jobs people business change - texas a&m university · sector’s declining importance prompt...

The Face of Texas

FEDERAL RESERVE BANK OF DALLAS • OCTOBER 2005

Jobs

People

Business

Change

The only constant is change itself. A Greek philosopher wrote those

words more than two millennia ago, at a time when events probably didn’t

move as rapidly as they do today. Accepting the inevitability of change still

leaves us the task of understanding the often bewildering world around us.

Most of us need help in deciphering the meaning of the changes we can see

and identifying the changes we can’t see. That is the purpose of this collection

of essays—to give our readers solid, useful perspectives that will assist in

understanding the changing Texas economy.

The new century finds Texas in transition—no longer booming on

high oil prices, finally rebounding from the late 1990s technology bust and

still looking for the next economic driver. Our times are complex. Texas’

economy is being reshaped not only by what’s happening in the state and

nation but also by a globalizing world, an important part of which lies just

over the Rio Grande.

These essays explain the recent past, give a textured picture of the

current landscape and offer a glimpse of what changes may be over the

horizon. Our goal is straightforward: We want to promote a fuller

understanding of Texas’ evolving economy, so the state’s citizens will be better

equipped to take advantage of opportunities that lie ahead.

Harvey Rosenblum

Executive Vice President and Director of Research

Federal Reserve Bank of Dallas

1 IntroductionMine Yücel

3 Texas in the Most RecentRecession and RecoveryMine Yücel

6 Why Did Texas Have a Jobless Recovery?Pia M. Orrenius, Jason L. Saving and Priscilla Caputo

12 Industry Clusters in TexasLaila Assanie and Mine Yücel

17 Economic Progress in the Texas EconomyRobert W. Gilmer

24 Texas Border Benefits from Retail Sales to Mexican NationalsKeith R. Phillips and Roberto Coronado

27 Texas Border Employmentand Maquiladora GrowthJesus Cañas, Roberto Coronadoand Robert W. Gilmer

33 Do Higher Oil Prices StillBenefit Texas?Stephen P. A. Brown andMine Yücel

37 The Changing Face of Texas:Population Projectionsand ImplicationsD’Ann Petersen and Laila Assanie

Table of ContentsThe Face of TexasJobs, People, Business, Change

1OCTOBER 2005 | FEDERAL RESERVE BANK OF DALLAS

T

IntroductionMine Yücel

The economic landscape of Texas ischanging. The state lost more than200,000 jobs during the tech bust andrecent recession. A majority of these jobswere in the high-tech sector, which wasthe main driver of the Texas economy inthe 1990s. With the decimation of the techsector, which industry will be the driver ofthe Texas economy in the future? Thismonograph doesn’t attempt to answerthat question, but we explore some of theways the Texas economy has been chang-ing and some of the current issues facingit.

The state has gone through boom andbust cycles before, but each downturn hasbeen followed by a stronger and morediverse economy. The oil bust was partic-ular to Texas and hurt the state’s economy,while low oil prices helped the rest of thenation. The tech bust, on the other hand,was experienced similarly in Texas andthe nation. Texas bore a larger bruntbecause it had a higher share of high-techmanufacturing and service industriesthan the nation.

While the drivers of the economy maychange, one constant is the close relation-ship the state has with the Mexican econ-omy. The interconnection is crucial to theborder economies and is a big factor inthe changing demographics of Texas.

Structural ChangeThe articles in this publication dis-

cuss some of the changes in the economiclandscape of our state. Mine Yücel looks atthe Texas economy’s performance duringthe most recent recession and explainswhy it was different from previous reces-sions. She argues that unlike previousrecessions, the most recent recession wasprimarily due to a high-tech bust ratherthan an oil price shock. Although oilprices were relatively high during therecession, they did not benefit Texas asmuch as in the past because the state hasdiversified away from oil. In addition, sheshows that the high-tech sector grew veryfast in Texas in the 1990s, to a share higherthan the national average. Texas’ highershare of industries that were hit hard in

this recession was a major factor in thestate’s prolonged downturn.

Pia Orrenius, Jason Saving andPriscilla Caputo survey the weak joblessrecovery after the most recent recessionand suggest that it may be caused bystructural change in the Texas labor mar-ket. They note that structural change isnot new to Texas. The state went throughstructural change in the 1980s after the oilbust and may be going through anotherone now. They show that the high-techand apparel industries are undergoingstructural losses, while the health care,education and government sectors areundergoing structural gains. But, just asthe oil industry decline paved the way forthe diversification and growth of the Texaseconomy a decade later, the structuralchange going on today will pave the wayfor a more dynamic and prosperous Texas.

Oil’s Impact The oil industry has been undergoing

change for the past 20 years, shrinkingwhile other sectors of the Texas economyhave grown. The Texas economy’s diversi-fication away from energy and the energysector’s declining importance promptStephen Brown and Mine Yücel to askwhether high oil prices are still a benefit tothe Texas economy. They show that higherenergy prices still benefit the state—eventhough it is by less than in the boom yearsof the 1970s and early ’80s. They also findevidence that the Texas economy hasbecome less sensitive to fluctuations in oilprices than it was in the ’70s and ’80s.First, oilfield activity has become less sen-sitive to fluctuations in energy prices. Sec-ond, the energy industry makes up asmaller share of the Texas economy than itused to. Together these factors mean thatTexas output is about 15 percent as sensi-tive to oil price fluctuations as it was from1970 to 1988. Texas employment nolonger seems to be positively affected byoil price fluctuations.

Business MixLaila Assanie and Mine Yücel outline

the importance of industry agglomerationto an economic growth. They highlight thekey clusters in Texas and its six major met-ropolitan areas through economic baseanalysis. They find that oil and gas extrac-tion and its support activities, pipelines,natural gas distribution, refining and oil-

2 FEDERAL RESERVE BANK OF DALLAS | OCTOBER 2005

field-machinery manufacturing are stillagglomerated in Texas. However, high-tech and transportation industries havebeen added to this mix. Computers,telecommunication services, semicon-ductors and air transportation firms nowhave a larger presence in Texas than in thenation as a whole.

Bill Gilmer analyzes per capita in-come growth in various regions of theTexas economy. He shows that the stateeconomy has been growing rapidly since1969, either matching or exceeding thenation’s growth. But the Texas Trianglecities of Houston, Dallas/Fort Worth,Austin and San Antonio grew faster thanaverage. Outside the Texas Triangle,income growth was much slower,although population growth was not.Especially after 1989, the Texas Trianglecities contributed three-fourths of thestate’s income growth. Gilmer notes thatthe Mexican border area represents achallenge to state economic developmentbecause the border cities’ average percapita income is only 50 to 60 percent ofthe national average. The border sawexplosive gains in the ’90s following thepassage of NAFTA and the growth of themaquiladora industry, but high popula-tion growth and high in-migration rateskept income per capita low in this area.The article also explains that the growth inwages and salaries after 1989 camethrough a change in industry mix as theeconomy shed low-wage jobs andreplaced them with better-paying ones.

Border InfluenceTexas border cities are a unique blend

of U.S. and Mexican cultures, languagesand customs and follow the ups anddowns of the Mexican and U.S.economies. Keith Phillips and RobertoCoronado look at how border cities on theTexas side benefit from cross-border traf-fic by consumers from their sister cities onthe Mexican side. They estimate retailsales in four metro areas along theTexas–Mexico border. They find that in2001, retail sales to Mexican nationalsaccounted for nearly 20 percent of retailsales in border metros. Laredo had thehighest share, with 41 percent of its retailsales going to consumers from across theborder. Phillips and Coronado also showthat unexpected changes in the peso’s realvalue affected these border metros

because retail sales strength varied closelywith peso strength, especially in Laredoand McAllen.

Another perspective on border citiesis presented by Jesus Cañas, RobertoCoronado and Bill Gilmer. They show howexpansions and contractions of the ma-quiladora industry have affected Texasborder cities. NAFTA’s passage and thepeso devaluation in the early ’90s led tomaquiladora growth and the relocation ofcomponent parts and material suppliersto Texas cities along the border. Texas bor-der cities developed rapidly in the ’90s aspart of this supply chain. Cañas, Coron-ado and Gilmer observe that proximity tothe U.S. market becomes a crucial advan-tage for the maquiladoras when there is ashort inventory cycle, when the weight-to-value ratio of goods is high, when thereis frequent retooling, when quality is moreimportant than price and when intellec-tual property rights are critical. However,they note that the state is unlikely torepeat the banner performance of the ’90sas foreign competition slows the growthof Texas border-city suppliers to themaquiladora industry. Hence, Texas maysee less stimulus in the future frommaquiladora expansion.

Population ShiftFinally, D’Ann Petersen and Laila

Assanie discuss Texas’ changing demo-graphic makeup and how it will shape theeconomy. Texas’ population is fastergrowing, younger and more ethnicallydiverse than the nation’s. The Hispanicpopulation will be the dominant force inTexas by 2020. The authors demonstratethat there are large disparities betweenethnic groups in income and education.Such disparities may imply a decline inreal income and a lower-skilled and less-educated labor force in Texas comparedwith the nation. On the other hand, theyoung and fast-growing population alsomeans Texas’ housing market may con-tinue to be vibrant even as the babyboomers age. Texas’ challenge is to reducethe disparities and make our differenceswork for us.

Yücel is a senior economist and vice presi-dent in the Research Department of theFederal Reserve Bank of Dallas.

Texas’ population is

faster growing,

younger and more

ethnically diverse

than the nation’s.

The Hispanic

population will

be the dominant

force in Texas

by 2020.


AAfter decades of faring better than therest of the country, Texas’ economicgrowth has lagged both the nation’s andits own past performance for almost threeyears.

The most recent U.S. recession wasshort-lived, beginning in March 2001 andending that November, according to theNational Bureau of Economic Research. Ittook Texas another 20 months—until July2003—to bottom out, based on the TexasCoincident Index.1 Employment growth

picked up in Texas in 2004. But while the1.7 percent increase put Texas on par withthe nation, it still left the state below itshistorical pace. What are the reasons forTexas’ prolonged downturn? Why did thestate lose its edge?

Past PerformanceTexas employment growth, on aver-

age, exceeded the nation’s from 1970through 2004, with a 2.8 percent rate tothe country’s 1.8 percent (Chart 1). The

Texas in the Most Recent Recession and RecoveryMine Yücel

Chart 1Texas Bests U.S. Employment Growth for More Than 30 Years

Percent

–2

–1

0

1

2

3

4

5

6

7

8

United States

Texas

’05’03’01’99’97’95’93’91’89’87’85’83’81’79’77’75’73’71

U.S. average

Texas average

NOTES: Data are year-over-year, seasonally adjusted, annualized rates. Shaded bars signify national recessions.

SOURCES: Bureau of Labor Statistics; Federal Reserve Bank of Dallas; National Bureau of Economic Research.

What are the

reasons for

Texas’ prolonged

downturn? Why

did the state

lose its edge?


sector losing 280,000 jobs nationwidefrom March 2001 to July 2003. These joblosses plus those in high tech constituted50 percent of the total U.S. employmentdecline. Texas lost 21,200 transportationjobs, which, combined with its high-techlosses, accounted for 62 percent of thestate’s total.2

Chart 5 illustrates Texas’ high-techroller coaster. California is included as acomparison, along with the United States.High tech grew very fast in the 1990s butcame back down just as fast. High-techproduction in Texas grew six times as fastas the state’s overall output. During therecession, Texas high-tech manufacturinglost 107,400 jobs, nearly a third of itsemployment. Even though Californiastarted with a higher base and thereforegrew less in percentage terms, more jobswere created in Texas. During the build-up, total high-tech manufacturing jobsincreased by 47,000 in Texas, while theyrose by only 17,000 in California. In semi-conductors, for example, California added22,000 jobs, while Texas added 35,000.Texas also grew faster than the nation intelecom services, adding 50,000 jobs dur-ing the ’90s, then losing 30,000 during therecession.

Elsewhere in this publication, “DoHigher Oil Prices Still Benefit Texas?” dis-cusses how the relationship between oiland the Texas economy has evolved.When the industry was a larger share ofthe Texas economy, higher oil prices werealways a net benefit to the state. That

state’s ability to dodge national recessionsis one reason Texas has done so much bet-ter. Eight of the 10 post–World War IIrecessions followed oil price shocks. Andunlike the nation as a whole, Texas bene-fited from high oil prices, especially in the1970s through early 1990s. As can be seenin Chart 2, the Texas economy followedchanges in oil prices fairly closely, withemployment rising and falling with the oilprice. The Texas employment cycle starteddiverging from oil-price movements inthe 1990s as the economy diversified awayfrom oil and gas.

High oil prices were a boon to theTexas economy and helped it grow, evenduring national recessions, as seen inChart 3. Oil prices that nearly tripled from$4 to above $10 per barrel (refiners’ acqui-sition cost) sent the United States intorecession in December 1973 but boostedoutput and employment in Texas (Chart3a). Oil prices started creeping up againin the late 1970s and rose from around $12per barrel in 1978 to almost $30 when Iraqinvaded Iran in September 1980. The U.S.economy went into recession — again,without Texas (Chart 3b).

But just as high oil prices helpedTexas, low ones hurt it. The nation wentinto recession again in August 1981, andTexas followed 10 months later, the resultof oil prices that began falling from recordhighs in March 1982 and the pull of thenational downturn (Chart 3c).

Texas had its own recession in 1986,when oil prices collapsed and the real

estate boom cratered. Low oil prices ben-efited the national economy but sentTexas into a steep decline. However, Texasskirted the national recession again in1990, when West Texas Intermediatecrude spiked to $45 per barrel with theIraqi invasion of Kuwait (Chart 3d).

This Time AroundTexas looked much more like the

nation in the 2001 recession than it did inpast downturns, for two reasons (Chart 4).First, although oil prices were high, thisrecession was primarily due to a high-techbust, not an oil price shock. Second, highoil prices do not help the Texas economyas much as they have in the past.

The collapse of high tech in therecent recession was greatly felt in Texas.The state had a larger share of high-techemployment than the U.S. average, so joblosses in those industries were relativelyhigher. From March 2001 through July2003 (the Texas recession), 39 percent ofthe jobs lost nationwide were in hightech—426,800 of them in manufacturingand 610,000 in services. Fifty-one percentof the 208,900 jobs lost in Texas were inhigh tech—51,900 of them in manufac-turing and 55,500 in services.

The events of September 11 also con-tributed to Texas’ steep downturn. Thetransportation industry is important tothe state’s economy and has a larger shareof total employment than in the nation.Transportation was especially hard-hit byfallout from the terrorist attacks, with the

Chart 2Texas Economy Follows Oil Prices

Detrended Texas employment 2005 U.S. dollars per barrel

Employment cycles

Oil price

–6

–4

–2

0

2

4

6

8

’05’03’01’99’97’95’93’91’89’87’85’83’81’79’77’75’730

10

20

30

40

50

60

70

80

90

NOTE: Employment data are seasonally adjusted and have had the time trend removed.

SOURCES: Bureau of Labor Statistics; Federal Reserve Bank of Dallas; Department of Energy.

After oil prices crashed,

Texas diversified and

the industry became a

much smaller share of

the state’s economy.


changed in the late 1980s, when volatileenergy prices helped erode the promi-nence of energy-intensive and energy-producing industries.

After oil prices crashed, Texas diversi-fied and the industry became a muchsmaller share of the state’s economy. Forexample, oil and gas output, whichaccounted for nearly 20 percent of totalTexas output in 1981, accounts for onlyabout 6 percent today. Similarly, oil andgas jobs account for only 2 percent ofTexas employment, down from a high ofabout 5 percent in 1982. The upshot isthat rising oil prices benefit Texas muchless now than they did in the past.

Texas is still a large producer andexporter of oil and gas, and when pricesgo up, it helps producers, royalty ownersand the state through increased severancetaxes. So, unlike the rest of the country,Texas gets an offset. But that offset ismuch less now than it was 25 years ago.

In sum, Texas’ economic perform-ance has been below par the past threeyears. Unlike other downturns, the 2001recession was primarily due to a high-techbust, not an oil price shock. And althoughoil prices were relatively high, they did notbenefit Texas as much as in the pastbecause the state economy has diversi-fied. In addition, high tech grew very fastin Texas in the 1990s, to a share that washigher than the national average. Texas’higher share of industries that were hard-hit in the recent recession was a majorfactor in the state’s prolonged downturn.

Yücel is a senior economist and vice presi-dent in the Research Department of theFederal Reserve Bank of Dallas.

Notes1 The Texas Coincident Index aggregates the

movements of key regional indicators—employment growth, the unemployment rateand gross state product—to gauge the state’soverall economic direction.

2 One point to note is that both high-tech andtransportation employment were falling evenbefore the onset of the recession. The two sec-tors were responsible for 73 percent of all joblosses in Texas from December 2000 to July2003.

Chart 5The Boom and Bust of High-Tech Manufacturing…

Jobs: Index, January 1990 = 100

Texas

United States

’05’03’01’99’97’95’93’91 ’04’02’00’98’96’94’92’90

California

60

80

100

120

140

160

…and Telecom Services

Jobs: Index, January 1990 = 100

Texas

United States

70

90

110

130

150

170

190

210

’05’03’01’99’97’95’93’91 ’04’02’00’98’96’94’92’90

California

Chart 4Texas Looks More Like the Nation(Total nonfarm employment)

Index, April 2001 = 100

Texas

United States

U.S. recessionApril – November 2001

97

98

99

100

101

102

Dec.’04

July’04

Feb.’04

Sep.’03

Apr.’03

Nov.’02

June’02

Jan.’02

Aug.’01

Mar.’01

Oct.’00


SOURCES: Bureau of Labor Statistics; Federal Reserve Bank ofDallas.

Chart 3Texas and U.S. Economies in Previous Recessions

a. Texas Skirts Recession in the 1970s…(Total nonfarm employment)

Index, December 1973 = 100

94

96

98

100

102

104

106

108

110

Texas

United States

Sept.’75

June’75

Mar.’75

Dec.’74

Sept.’74

June’74

Mar.’74

Dec.’73

Sept’73

June’73

U.S. recessionDecember 1973 – March 1975

b. …and Again in 1980(Total nonfarm employment)

Index, February 1980 = 100

94

96

98

100

102

104

106

108

110

Texas

United States

June’81

Apr.’81

Feb.’81

Dec.’80

Oct.’80

Aug.’80

June’80

Apr.’80

Feb.’80

Dec.’79

Oct.’79

Aug.’79

U.S.recession

February –July 1980

c. Texas Follows Nation into Recession in 1981…(Total nonfarm employment)

Index, August 1981 = 100

94

96

98

100

102

104

106

108

110

Texas

United States

Nov’83

Aug.’83

May’83

Feb.’83

Nov.’82

Aug.’82

May’82

Feb.’82

Nov.’81

Aug.’81

May’81

Feb.’81

U.S. recessionAugust 1981 – November 1982

d. …but Dodges the 1990 Downturn(Total nonfarm employment)

Index, August 1990 = 100

Texas

United StatesU.S.recession

August 1990 – March 1991

94

96

98

100

102

104

Feb.’92

Nov.’91

Aug.’91

May’91

Feb.’91

Nov.’90

Aug.’90

May’90

Feb.’90



I

Why Did Texas Have a Jobless Recovery?Pia M. Orrenius, Jason L. Saving and Priscilla Caputo

ers on firm payrolls. However, the house-hold survey has a much smaller samplesize and depends on population estimatesthat are not always reliable, mainlybecause of uncertainty about immigra-tion rates. Given these weaknesses andthe adoption of a statistical method tocompensate for missed job growth instart-up firms, most experts—and theBLS—consider the payroll survey the bet-ter gauge of employment.

Productivity Growthor Uncertainty?

If the data are sound and the countrydid experience a jobless recovery in 2002and 2003, could high productivity growthor substantial uncertainty have been thecause?

U.S. productivity growth averaged 4.3percent during this period, and someexperts believe that increase—well abovethe post–World War II average of about 2 percent—enabled companies to step up production without hiring more work-ers. Others believe the uncertain environ-ment that followed various corporateaccounting scandals and the 9/11 attacksled to a wait-and-see approach by em-ployers.

In early 2001, the U.S. and Texaseconomies fell into recession. While theNational Bureau of Economic ResearchBusiness Cycle Dating Committeedeclared the U.S. recession over inNovember of that year, job growth did notresume until June 2003. Texas job growthbroke into positive territory two monthslater, and there is evidence that, like thenation, economic activity in the statepicked up long before that.

Following a typical recession,employment begins to rise at about thesame time output does. But in the twoyears after the 2001 recession, U.S. realoutput growth averaged 2.5 percent, whileemployment growth was essentially zero.The divergence between output andemployment was even more pronouncedin Texas, where real output—as measuredby gross state product—grew faster thanthe nation’s, but employment fell at an average annual rate of 0.2 percent. Clearly, something was different this time.

Many explanations have been offeredfor the unusually weak labor market per-formance, including problems with meas-uring employment, high productivitygrowth, widespread uncertainty in thewake of 9/11 and corporate scandals, and

structural change in the economy. Whilemuch has been written on the nation’sexperience during this period, there is lit-tle information on what caused the job-less recovery in Texas. For this reason, it’simportant to examine these explanationsto see which of them can shed light on thestate’s experience.

Employment Statistics?Two Bureau of Labor Statistics (BLS)

surveys are the primary source fornational and state employment data. Theestablishment, or payroll, measure—offi-cially, Current Employment Statistics—surveys about 400,000 work sites eachmonth. Critics contend this survey under-states job creation at economic turningpoints because it misses employment inthe new firms created during a recovery’sinitial stages. The alternative, household-based Current Population Survey contactsindividuals directly about their employ-ment status. According to this survey,there has been little jobless about therecovery: Jobs have grown each year sincethe 2001 recession.1

The household survey might seemsounder than the payroll survey becauseit is not limited to wage and salary work-


These factors likely played an impor-tant role in the jobless recovery. But jobgrowth in 2002–03 was far below whatTexas and the nation saw in earlier peri-ods of relatively high productivity growth,such as the late 1990s, and substantialuncertainty, such as the late 1970s. Sothere is more to the story.

Structural Change? A widely read article from the Federal

Reserve Bank of New York offers anotherexplanation for the jobless recovery.2 EricaGroshen and Simon Potter consider twotypes of effects that could shake up labormarkets: (1) short-term cyclical adjust-ments that vary with the business cycle,and (2) longer term structural changes, inwhich some industries decline while oth-ers grow.

The economists contend that anunusually large amount of structuralchange in the labor market, as opposed totemporary cyclical adjustments, hinderedthe resumption of employment growth in2002 and 2003. When jobs shift acrossindustries, new positions have to be cre-ated and filled, which takes far more timethan simply recalling workers to theirjobs, as might occur with cyclical change.So if structural change is on the rise, itcould explain the jobless recovery.

The kind of structural changeGroshen and Potter consider can resultfrom a myriad of factors that cause someindustries to decline as others grow. Thesefactors include technological and demo-graphic change, reorganization of pro-duction, trade and outsourcing—any oneof which can permanently alter a state’s ornation’s industrial mix. Cyclical job losses,by contrast, move with the business cycle.As the economy enters a recession, jobsare temporarily lost in response to soften-ing demand. They are added back as theeconomy picks up again.

Looking at job growth by industry,Groshen and Potter find that structuralfactors played a much greater role in theUnited States during 2001–02 than in earlier U.S. recoveries. They attribute thisto a changing labor market in which cycli-cal job losses have been minimized andstructural changes are more pervasive.

This conclusion has important im-plications for public policy. The tradi-tional safety net in the United States, withsuch elements as unemployment insur-

ance, is largely designed around the needsof the cyclically unemployed — peoplewho need short-term help with incomesustenance while they search for a job.The system is generally not designed toprovide longer term retraining for dis-placed workers whose sectors perma-nently shrink. Public job-training pro-grams are becoming more common, how-ever. Lawmakers recognized the effects ofstructural change in the labor market inpassing such bills as the Workforce Invest-ment Act of 1998 and the Trade Adjust-ment Assistance Reform Act of 2002.

Assuming structural change hasaccelerated at the national level, can thesame be said for Texas? Taking theGroshen–Potter approach, we comparerecent patterns to earlier recessions to seeif structural change has increased in Texasand, if so, whether it helps explain thestate’s recent experience.

Measuring Structural Change To measure structural job change,

Groshen and Potter compare employ-ment growth in the recession and therecovery.3 They make this comparison foreach major industry over the length of therecession as designated by the NationalBureau of Economic Research (NBER)—March 2001 to November 2001.4 The recov-ery is defined as the 12 months followingthe business cycle’s trough in November.

Pinpointing recession dates for Texasis more complicated. Economic analystsoften look to payroll employment growth todate state recessions because this is themost timely and reliable data available at

the state level. In a jobless recovery, how-ever, the traditional relationship betweenemployment growth and overall eco-nomic activity breaks down. This meanspayroll employment data may not haveaccurately reflected the state’s overall eco-nomic health during 2002–03, making itimpossible to date the Texas recessionusing those numbers.

The NBER solves this conundrum forthe nation by using several variables inaddition to employment—such as indus-trial production and, especially, grossdomestic product—to date U.S. businesscycles.5 Most of these numbers are notavailable in a timely fashion at the statelevel, and they are not available at all on aquarterly or monthly basis, which wouldbe needed to date the Texas recession.

We use the national dates for a base-line analysis of Texas. After all, Texas em-ployment closely tracked the nation’s in2001 and thereafter, suggesting that simi-lar factors drove both economies intorecession (Chart 1 ). Texas output alsotracked the nation’s reasonably well in2001 and 2002 (Chart 2). That said, esti-mates of real output at the state level aresubject to a higher degree of uncertaintythan at the national level, and there isanecdotal evidence Texas emerged fromthe recession after the nation. To checkthe validity of our findings, we repeat theanalysis using an end date of March 2003rather than November 2001.6

Chart 3 shows how Texas job growthfared during the 2001 recession and the12-month recovery for each one-digitindustry, the broadest category in the

Chart 1Texas Employment Tracks the Nation’s in 2001 and After

Index, 2000 = 100

Texas

40

50

60

70

80

90

100

110

’04’02’00’98’96’94’92’90’88’86’84’82’80’78

United States

NOTE: Data show total nonfarm employment.

SOURCES: Bureau of Labor Statistics; Federal Reserve Bank of Dallas.


Standard Industrial Classification (SIC)system.7 All growth rates are relative to theaverage for total Texas employment dur-ing the relevant period. For example, if anindustry grew 5 pecent slower than theTexas economy as a whole during therecession, its growth rate is –5 percent.Likewise, if an industry grew 5 percentfaster than the Texas economy during therecovery, its growth rate is 5 percent.

The horizontal axis on Chart 3 measures the relative growth rate duringthe recession; the vertical axis measuresthe relative growth rate during the recov-

ery. If an industry grew slower than thestatewide average in each of the two peri-ods, it falls in the southwest portion of thechart, labeled “structural losses” becausethese industries lose jobs regardless ofoverall economic conditions. If an indus-try grew more rapidly than the statewideaverage during both intervals, it is in thenortheast portion of the chart, labeled“structural gains” because such industriesgain jobs regardless of the overall econ-omy.

The remaining quadrants deal withindustries that rise and fall with the busi-

ness cycle. Industries that grew slowerthan the statewide average during therecession but faster during the recoveryare in the procyclical flows quadrantbecause they move with changes in thebusiness cycle. Industries that grew fasterthan the statewide average during therecession but slower during the recoveryare in the countercyclical flows quadrantbecause they tend to add jobs when therest of the economy declines but lose jobswhen the rest of the economy does well.

The size of each industry’s bubble onthe chart represents its share of total Texasemployment in March 2001, when therecession began. The larger the bubble,the larger the industry’s share of the state’sworkforce at that time.

The results suggest that the recentbusiness cycle has been dominated bystructural gains and losses, as most majorindustries fall into the structural changequadrants in Chart 3. Manufacturing ofboth durable and nondurable goods suf-fered the largest structural losses,whereas health services and governmenthad the biggest structural gains. Overall,about 75 percent of March 2001 employ-ment was concentrated in industries thatsubsequently underwent structuralchange. The next section breaks downthese major industries to take a closer lookat job adjustments.

Industries with Structural Loss. Indus-tries in Chart 4 are classified according tosubsectors in the North American Indus-try Classification System (three-digitNAICS codes). The southwest portion ofChart 4 includes a number of high-techsectors, among them computer and elec-tronic product manufacturing (includessemiconductors); electrical equipment,appliance and component manufacturing;telecommunications; and Internet serviceproviders (ISPs), search portals and dataprocessing services. High tech’s presencein the structural loss quadrant is not sur-prising, since the 2001 recession kickedoff a prolonged retrenchment and re-structuring for the sector in Texas, aprocess from which the state has not fullyemerged.

Apparel manufacturing also falls inthe structural loss quadrant. In contrast tohigh tech, the apparel industry has beendeclining in the United States and Texasfor many years. Indeed, apparel experi-enced the largest job losses in percentage

Chart 3Structural Change Prevalent Among Texas Industries in 2001–02

Job growth in recovery (percent)

–5

–4

–3

–2

–1

0

1

2

3

4

–6 –5 –4 –3 –2 –1 0 1 2 3 4 5

Procyclical flows

Structural losses

Structural gains

Countercyclical flows

ConstructionHealth services

Wholesaletrade

FIRE

Other services

TCPUDurable manufacturing

Mining

Nondurablemanufacturing Retail

trade

Government

Job growth in recession (percent)

NOTES: TCPU is transportation, communications and public utilities. FIRE is finance, insurance and real estate. The recession is datedMarch to November 2001; the recovery is dated December 2001 to November 2002.


Chart 2Texas Real Output Tracks the Nation’s in 2001 and After

Index, 2000 = 100

Texas

40

50

60

70

80

90

100

110

United States

’04’02’00’98’96’94’92’90’88’86’84’82’80’78

NOTE: Data show real U.S. gross domestic product and Texas gross state product.

SOURCES: Bureau of Economic Analysis; Federal Reserve Bank of Dallas.


terms during both the recession and therecovery.

The northeast quadrant of Chart 4shows the industries that grew faster thantotal Texas employment during the reces-sion and recovery. This quadrant consistsmainly of sectors related to the provisionof health care and education, includinglocal government.

Given recent policy and demographicdevelopments, this trend is understand-able. Rapid advances in medical technol-ogy, coupled with an aging population,are producing an increased emphasis onhealth care, regardless of the business cycle.

The rise in the economic return toeducation, the burgeoning youth popula-tion and renewed public attention to edu-cational quality have produced an in-creased emphasis on education thatdoesn’t ebb and flow with economic con-ditions, either. Since local government isthe largest provider of K–12 education, it’snot surprising that employment in thissector rose during the recession, as well asthe recovery.

Countercyclical Industries. The south-east corner of Chart 4 consists mainly ofindustries in the energy sector. Risingenergy prices were a contributing factorto the 2001 recession.8 As home to a majorshare of the U.S. energy industry, Texasbenefits from high oil prices (although toa lesser extent than when the industryconstituted a much larger part of the state’s economy).

Since energy prices were higher dur-ing the 2001 recession than during the2002–03 recovery, it makes sense thatenergy is categorized as countercyclicalfor this period.9 Natural resource and min-ing industries in this quadrant include oiland gas extraction and mining supportactivities.

One notable countercyclical industrythat doesn’t fit into the natural resourcecategory is real estate. What high oil pricesdid for natural resource industries duringthe recession, low-interest loans likely didfor homebuyers.

Procyclical Dating. Despite expecta-tions of “normal” cyclical losses, few in-dustries fall into the procyclical categoryduring and after the 2001 recession. Thenorthwest quadrant of Chart 4 consists ofonly about 9 percent of total employment.Among the industries in this quadrant areretail, transportation-related sectors and

accommodations. It may be surprising that so few in-

dustries fall into the cyclical category, butit’s important to remember that we arecomparing each industry to the overallstate economy. If an industry’s employ-ment fell slightly during the recession androse slightly during the recovery, it’s cate-gorized as countercyclical because itsemployment fell by less than the stateaverage during the recession and rose byless than the average during the recovery.

Recession Dating. What if the Texasrecession was longer than the nation’s anddid not end in November 2001? If so, theanalysis so far biases the findings towardstructural change by attributing 2002 joblosses to the recovery instead of to whatmay have been a continuing recession. Tocheck our results, we repeat the exerciseunder the assumption that Texas emergedfrom the recession in March 2003—muchlater than the nation and about fourmonths before employment growthresumed in the state.

A few industries move from one quad-rant to another, but the overall picture isone in which structural change still domi-nates cyclical change (Chart 5). About two-thirds of employment is concentrated inindustries undergoing structural change,compared with three-fourths when No-vember 2001 is used as the end date.

Chart 4Structural Change Prevalent in 2001–02: A Look at More Detailed Industries


–25

–20

–15

–10

-5

0

5

10

15

20

–20 –15 –10 –5 0 5 10

Procyclical flows

Structural losses

Structural gains


Warehousing &storage

Waste management

Funds, trusts& otherfinancialvehicles

Mining support activities

Pipelinetransportation

Telecommunications

ISPsElectricalequipment,appliance &component mfg.

Computer & electronicproduct mfg.

Apparel mfg.

Clothingstores

Railtransportation

Transportationequipment

Accommodations

Airtransportation

Real estate

Local govt.

Ambulatoryhealth services

Oil & gasextraction


NOTES: The recession is dated March to November 2001; the recovery is dated December 2001 to November 2002. Educational services isbehind local government in the northeast quadrant.


Comparing Texas RecessionsIs structural change a bigger factor

today than in the past? Groshen and Potter conclude that for

the United States as a whole, it is. More in-dustries in the recent recession fell into thestructural-change quadrants, comparedwith earlier recessions. They find that 79percent of U.S. employment was in indus-tries affected more by structural thancyclical shifts in the 2001 recession, upfrom about 50 percent in previous down-turns.

It’s a somewhat different story forTexas. Chart 6 shows job adjustments bymajor industry during the recession andrecovery of the early 1980s.10 That reces-sion was more severe than the recent one,with several large industries — such asdurable manufacturing and mining—experiencing double-digit job losses. Nev-ertheless, except for government, educa-tion and health services, the losses werefairly concentrated in the structural cate-gories. In fact, Texans were about as likelyto work in structural-change sectors in the1982–83 recession as they were in 2001.The share of structural job losses wasabout 72 percent during the earlierperiod, compared with 76 percent in the2001 recession. While the relationship canbe seen a bit more easily in Chart 3 than inChart 6, the two graphs confirm that


structural change, as defined by Groshenand Potter, is not new to Texas.

More Sectors UndergoStructural Change

Several explanations have beenoffered for the growing role of structuralchange in the U.S. economy, includingtechnological change and increasing

international trade. A decline in the role ofcyclical change, meanwhile, has beenlinked to factors such as improved mone-tary policy, which appears to have less-ened the duration and severity of U.S.business cycles.11 Better supply chain man-agement has also allowed firms torespond more quickly to changes indemand and avoid sudden large swings in

inventory, production and employment.Additional contributing factors are adecline over time in the severity of energyand food supply shocks and the deregula-tion of financial markets.

But does structural change reallyexplain the jobless recovery? Structuralchange, as measured here, was about asprevalent in Texas in the 2001 recessionand ensuing recovery as in the early 1980srecession and recovery. The differencebetween the two periods is the severity ofthe change. Job losses were much deeperin the 1982–83 downturn (and worse yetin 1986). Nevertheless, employmentrebounded with a short lag, and there wasnothing like the jobless recovery experi-enced post-2001.

Another possibility is that the invest-ment bust that characterized the 2001recession and its aftermath may havedriven both structural losses in the labormarket and the jobless recovery. Theinvestment bust followed the investmentboom that had characterized certain fast-growing industries—led by high tech—during the 1990s. In Texas, for example,post-2001 venture capital commitmentsfell sharply to about 20 percent of their2000 levels. The investment bust likelydelayed employment growth during therecovery in the sectors that had beenbooming. If this was the case, sectors thatwere fast-growing before the recessionwould fall into the structural loss categoryin our analysis. These industries may ormay not belong there, depending onwhether they will eventually resumeabove-average job growth.

The data suggest that the investmentbust played an important role in Texasduring the recent business cycle. In fact,of the state’s 16 fastest-growing industriesin the 1990s, 10 appear in the structural-loss quadrant of Chart 4, meaning theyshed jobs both during and after the 2001recession. Groshen and Potter show thatfor the nation, seven of the 18 fastest-growing industries fall into the structuralloss category.

It is likely that as these industries’expansion fell short of expectations,investment dried up and employmentdeclined. The industries include severalhigh-tech subsectors, such as telecom-munications and ISPs, search portals anddata processing services. Not all fast-growing industries fall into the structural

Chart 6Structural Change a Major Factor in 1982–84


–6

–4

–2

0

2

4

6

8

–20 –15 –10 –5 0 5 10 15

Procyclical flows

Structural losses

Structural gains


GovernmentHealth services

Retail trade

Educationalservices

Construction

TCPUMining

FIRE

Nondurable mfg.

Durable mfg.

Wholesaletrade

Other services


NOTES: TCPU is transportation, communications and public utilities. FIRE is finance, insurance and real estate. The recession is datedMarch 1982 to March 1983; the recovery is dated April 1983 to March 1984.


Chart 5Structural Change Prevalent with Alternate Recession End Date


–25

–20

–15

–10

–5

0

5

10

15

–50 –40 -30 –20 –10 0 10 20 30Job growth in recession (percent)

Procyclical flows

Structural losses

Structural gains


Apparelmfg.

Warehousing

Ambulatoryhealth care

Highway, street &bridge construction

MiningsupportactivitiesElectronics &

appliance stores

Telecommunications Pipelinetransportation

Computer &electronicproduct mfg.

State government

Local government

NOTES: The recession is dated March 2001 to March 2003; the recovery is dated April 2003 to March 2004.



loss quadrant, however. Three of Texas’fastest-growing industries in the 1990s arein the structural gain category—ware-housing and storage, ambulatory healthservices and social assistance.

Little New AboutStructural Change

The Texas economy has undergonefundamental restructuring as the statehas diversified away from agriculture andenergy and become more like the nation.These trends began in earnest in the 1970sand intensified in the 1980s with the dropin oil prices and collapse of the bankingsector. The 1990s saw tremendous growthof the state’s high-tech industries and fur-ther consolidation in the energy sector. Inboth the 1970s and again in the 1990s,Texas’ economic growth was character-ized by large inflows of workers whobrought different skills and educationwith them and contributed to the state’seconomic transformation.

The decline of industries paves theway for diversification and the growth ofnew sectors as workers, capital and know-how are freed up to pursue better ends.For example, at one time, 90 percent of theU.S. labor force was engaged in farming.Today, that number is a mere 3 percent.

While this transformation is clearlybeneficial in the long run, in the short tomedium term, this type of change is notwithout its critics. People may primarilysee the negative connotations of struc-tural change, without seeing the benefits.This may be because certain advances intrade and technological change have largebenefits that are spread across many peo-ple, such as all U.S. consumers, while thecosts of such advances can be small butconcentrated on a select few (such as laid-off textile workers).

Texas has not been immune to theforces of trade and outsourcing. Semicon-ductor production has moved out ofAustin and Dallas to Asia, for example,and major computer companies haveconcentrated their software developmentin India, outsourcing thousands of jobsthere. Big retailers and national bankscontinue to expand in the state, often dis-placing or absorbing local businesses inthe process.

At the same time, the state’s economyhas many strengths. Workers andinvestors continue to flock to Texas, home

construction is at record levels, freed-upcapital and labor are moving into sec-tors—such as education and health—where structural growth is most pro-nounced. Exports to China are booming,and the border economy is thriving as aresult of freer trade with Mexico.

SummaryThe aftermath of the 2001 recession is

often described as a jobless recovery. Ittook Texas and U.S. employment almostfour years to reach their respective prere-cession levels, which they finally did inJanuary 2005. Many factors contributed tolabor market weakness in 2002 and early2003, including high productivity growth,the war on terror and corporate scandals.

In their New York Fed article, Groshenand Potter highlight another potentialsource of labor market weakness—struc-tural change. The economists imply thatbecause new industries are replacing oldones, jobs are being created and filled at aslower rate than in past business cycles, inwhich workers were simply laid off andrehired by the same or similar employers.

Applying the Groshen–Potter meth-odology to Texas, we find that structuralchange also dominated cyclical change inthe state during the last business cycle. Wedo not find, however, that the amount ofstructural change has increased over time,as Groshen and Potter argue is the case forthe nation.

Structural change is an enduring fea-ture of the state’s economy. But whileTexas labor markets experienced struc-tural change in earlier recessions, they didnot experience drawn-out weakness oncea recovery was under way. In other words,the recent jobless recovery remains a bitof a mystery. The investment boom andsubsequent bust may have had somethingto do with it. Many of the 1990s’ fastest-growing industries ended up with thelargest relative and most persistent joblosses. The extent of the state’s high-techinvestment boom and subsequent bustmay help explain why the effect on Texasemployment growth was so significantand lasting.

Orrenius and Saving are senior economistsin the Research Department of the FederalReserve Bank of Dallas. Caputo worked onthis article while an economic analyst atthe Bank.

Notes1 Annual household employment was lower in 2002

than 2001, but yearly job growth is calculatedDecember-over-December and was 0.26 percentin 2002.

2 “Has Structural Change Contributed to a JoblessRecovery?” by Erica L. Groshen and Simon Pot-ter, Federal Reserve Bank of New York CurrentIssues in Economics and Finance, August 2003.

3 Alternative measures of structural change arediscussed at length in “Can Sectoral Realloca-tion Explain the Jobless Recovery?” by DanielAaronson, Ellen R. Rissman and Daniel G. Sulli-van, Federal Reserve Bank of Chicago EconomicPerspectives, Second Quarter 2004.

4 Groshen and Potter compare employmentgrowth in the recession and the recovery fortwo-digit industries as defined by the StandardIndustrial Classification (SIC) system. SIC codeswere replaced by the North American IndustryClassification System (NAICS) in 2002.

5 In a statement announcing the dating of the 2001 recession, the NBER called real GDP“the single best measure of ‘aggregate eco-nomic activity.’” See www.nber.com/cycles/recessions.pdf.

6 Texas payroll employment began to grow inAugust 2003, while retail sales began to grow inSeptember 2002. As a compromise, we selectedMarch 2003 as an alternative end date for thestate recession. Eleventh District Beige Bookaccounts also suggest the second quarter of2003 may have been the turning point for Texas.

7 SIC codes are used in Charts 3 and 6 so thatemployment by industry can be compared overtime. The newer, three-digit NAICS codes areused in Charts 4 and 5.

8 See “Do Energy Prices Threaten the Recovery?”by Stephen P. A. Brown, Federal Reserve Bank ofDallas Southwest Economy, May/June 2004.

9 In 2004, oil prices rose again, and they are cur-rently higher than they were during the 2001recession. Natural gas prices have alsoremained high.

10 The 1982–83 Texas recession is assumed tohave lasted from March 1982 to March 1983.This period roughly corresponds to the down-turn in both state output and employment.

11 See “New Economy, New Recession?” by EvanF. Koenig, Thomas F. Siems and Mark A. Wynne,Federal Reserve Bank of Dallas Southwest Econ-omy, March/April 2002.


A

Industry Clusters in TexasLaila Assanie and Mine Yücel

Advances in technology have dramat-ically reduced transportation and com-munication costs. Access to distant goods,services and even labor has become mucheasier. Increased access to markets hasalso brought increased competition, pres-suring firms to reduce costs to maintainprofitability. In this age of globalization, asMichael Porter notes, the importance ofgeneralized urban economies diminishes,and agglomeration economies becomemuch more important.1

An agglomeration economy, alsoknown as a cluster, is defined as a geo-graphically concentrated group of indus-tries related by technology or skills, withclose linkages among buyers and suppli-ers. Clusters are important because theyprovide their participants with easy andlucrative access to knowledge and special-ized resources required to operate effi-ciently. This enhances participants’ pro-ductivity and spurs innovation. Clustersalso attract new business and investment


to the region. It is this increased efficiencyand the ability to innovate and attract out-side investment that give cluster partici-pants a competitive advantage.2 A goodexample of an industry cluster is the Dal-las telecom corridor that attracted hun-dreds of high-tech manufacturing andservices firms to the metro during thehigh-tech boom in the 1990s.

Industry clusters lead and shape theeconomic growth of a region. One simpleway of determining industry clusters isthrough economic base analysis. The eco-nomic base of a region is defined asindustries whose external demand gener-ates outside revenues and stimulates localeconomic growth. The assumption is thatnontraded goods and services tend to beuniformly distributed, do not bring out-side income into the region, and there-fore, do not form the region’s economicbase.

To determine which goods and serv-ices produced in Texas and its major met-ropolitan areas are basic, or exportable,we use location quotients (LQ), a toolcommonly used to analyze the economicbase of a region. Location quotients com-pare the local economy with a referenceeconomy (for example, the Dallas econ-omy with the U.S. economy) to identifyareas of specialization. The quotients arecomputed as follows:

local employment in industry i/U.S. employment in industry i

total local employment/total U.S. employment

Location quotients higher than 1indicate that the regional concentrationof these industries is greater than theirnational concentration and so they arelikely to be part of the economic base ofthe region. The greater the location quo-tient, the higher the concentration andthe more certain we are of the basicnature of the industry.3

Our location quotients are computedusing 2000 census employment data fromIntegrated Public Use Microdata Series(IPUMS) files.4 We first analyze Texas’ eco-nomic base and compare the state’s geo-graphic dispersion of industries with thatof the nation.5 We then look at the basicactivities of the six major metropolitanareas in Texas and compare the degree of

agglomeration with both Texas and theUnited States.

How Texas ComparesWith the United States

Historically, Texas has been known foroil, cotton and cattle. But in recentdecades the state’s image has changedsubstantially. Today, the economic base ismore diverse and includes transportation,computer, semiconductor and telecom-munication firms.

Chart 1 plots the location quotients ofTexas with the United States as the refer-ence region.6 The chart shows that eventhough Texas has diversified, the energyindustry is still a large part of its economicbase. Oil and gas extraction and its sup-port activities, pipelines, natural gas dis-tribution, refining and oilfield-machinerymanufacturing are still agglomerated inTexas. However, high-tech and trans-portation industries have been added tothis mix. Computers, telecommunicationservices, semiconductors and air trans-portation firms now have a larger pres-ence in Texas than in the nation.

Looking first at energy-related indus-tries, the state’s share of employment inoil and gas extraction and mining is nearlysix times the national share. Much lessdramatic, yet significant, are the locationquotients for high-tech manufacturingand services. The share of computer andperipheral equipment manufacturing inTexas employment is 78 percent higherthan in the nation, wired telecommunica-tion services 50 percent higher, other tele-

com services 21 percent greater and elec-tronic components manufacturing, whichincludes semiconductor manufacturing,is 44 percent higher than in the nation.Finally, the nation’s employment share inair transportation is approximately 60percent less than that of Texas. Moreover,the high-tech, oil and gas, and air trans-portation industries are the largest em-ployers in Texas, confirming their impor-tance to the state’s economic base.

Since higher concentration indicatesthe presence of clusters, or specialization,these industries are the central drivers ofthe Texas economy. The prominence of these industries—high tech, telecom-munication services and air transporta-tion—helps to explain why the state’seconomy fared worse than the nation’sduring the recent downturn. Despite thehigh-tech bust, high-tech manufacturingand service sectors remain clustered inTexas.7

Based on location quotients, nearly 35percent of Texas’ employment is in indus-tries that can be classified as “basic” or ex-portable. In these basic industries, 17 per-cent of Texas’ workforce produces goodsand services that satisfy nonlocal demand.

Texas Major MetrosThe major metropolitan areas in

Texas account for more than two-thirds ofthe state’s employment, so the sectors thatlead these metro economies determinethe state’s overall economic base. Thecomposition of economic activity variessignificantly among Texas’ major metros.

Chart 1Texas Compared with the United States

0

1

2

3

4

5

6Support activitiesfor mining

Oil and gas extraction

Trucktransportation

Petroleum refining

Home health care services

Wired telecomcarriersAir transportation

Computer systemsdesign and related services

Chemicals mfg.

Pipeline transportation

Computer and peripheralequipment mfg.

Natural gasdistribution

Aircraftand partsmfg.

Electronic componentsand products mfg.

Communicationsequipment mfg.

Petroleum andproducts

Oilfieldmachinery mfg.

Services incidentalto transportationOther

telecomservices

Locationquotients

SOURCE: 2000 Census IPUMS data; authors’ calculations.

LQi =


Each metro area specializes in a uniqueset of industries, diversifying and strength-ening the state’s economy.8

Dallas. Location quotients for Dallasconfirm the metro is the high-tech mecca,transportation hub and telecommunica-tion nexus of Texas. Dallas’ share of work-ers in other telecom services, air trans-portation services, communicationsequipment manufacturing and computersystems design and services is twice ormore when compared with the overallstate. This concentration of high techbecomes even more striking when themetro is compared with the United States(Chart 2). Employment shares of commu-nications equipment manufacturing arefour times greater than in the nation,while those of telecom services (bothwired and other) and air transportationexceed the national shares by three times.Moreover, although Dallas doesn’t havemuch energy industry concentrationcompared with the state, oil and gas ex-traction is—surprisingly—Dallas’ fifth-most-concentrated industry in compari-son with the nation. Despite the recentdownturn, high-tech manufacturing andservices firms remain key contributors ofthe metro’s economic base.

Austin. High tech and state govern-ment compose Austin’s economic base.This is evident from the metro’s locationquotients, which surpass the state’semployment shares in computer andperipheral manufacturing (6.4 times),electronic components manufacturing(4.8 times), public administration of envi-ronmental quality programs (3.8 times),public finance activities (3.6 times), andexecutive and legislative bodies (3.3times). Compared with the nation,Austin’s larger presence of computer andchip makers is even more pronounced(Chart 3), and the metro’s share of work-ers employed in these industries signifi-cantly exceeds the national share (11.5and 6.9 times, respectively). Although therecent high-tech bust hit Austin hard,semiconductor and computer manufac-turing industries remain key elements ofthe metro’s economic base.

Houston. Houston is Texas’ oil and gascapital and home to the sixth largest sea-port in the world. Not surprisingly, thelocation quotients convey a similar story.Five of the 10 most geographically con-centrated industries compared with the

Chart 3Austin Compared with the United States


Locationquotients

0

2

4

6

8

10

12


Radio, TV andcomputer stores

Other informationservices

Data processingservices

Legal services Colleges anduniversities

Computer system designsand related services

Public financeactivities

Executive officesand legislative bodies

Adminstration ofhuman resourceprograms


Chart 4Houston Compared with the United States

Support activitiesfor mining

Locationquotients

0

2

4

6

8

10

12

14

Oil and gasextraction


Petroleumrefining

Computer andperipheral equipment mfg.

Airtransportation

Chemicals mfg.

Oilfield machinery mfg. Petroleum and petroleumproducts, wholesale

Architectural andengineering services

Watertransportation

Pipelinetransportation

Management ofcompanies andenterprises

Metals and minerals, exceptpetroleum, wholesale

Petroleum andcoal mfg.


Chart 2Dallas Compared with the United States

Communications, audio andvideo equipment mfg.

0

1

2

3

4

5


Structural clay products mfg.



Aircraft andparts mfg.

Railroadmfg.

Radio, TV andcomputer sales

Airtransportation

Management ofcompanies andenterprises

Data processingservices

Nondepositorycredit andrelatedactivities

Computersystemsdesigns andrelated services

Othertelecomservices

Wired telecomservices

Locationquotients



state as well as the nation are related to oiland gas production, drilling and oil serv-ices (Chart 4). Given that the oil and gasindustry is more prevalent in Texas than inthe United States, Houston exhibits muchstronger oil- and gas-related industryclusters in reference to the nation than toTexas. Location quotients for bothupstream and downstream oil and gasactivities such as extraction, supportactivities for mining, pipeline transporta-tion, petroleum refining and wholesalingof petroleum products more than doublewhen the base region changes from Texasto the United States. Downstream activi-ties such as petroleum refining and chem-icals manufacturing also bolster portactivity. Hence, the share of water trans-portation employment is twice as high inHouston when it is compared with the restof Texas (2.6 times) as well as the nation(2.4 times). Also, because Texas has a highshare of computer manufacturing, Hous-ton’s edge over the nation in computerand peripheral equipment manufacturing(2.5 times higher) is larger than its edgeover Texas (1.4 times higher).

San Antonio. San Antonio’s economicbase thrives on tourism and the presenceof large insurance firms, electric and gasproduction and distribution firms, andfour military bases. Also, the metro has asignificant presence of health care organi-zations and recently has become home toseveral telemarketing companies (Chart5). The metro exhibits similar employ-ment share ratios when compared withthe state or nation. First, San Antonio’sshare of employment in national securityand international affairs is more than fivetimes that of the nation as well as thestate, largely because of the strong mili-tary presence in the metro. Second, thereis sizable specialization of insuranceproviders and electric and gas producersand distributors. Third, San Antonio hasmore than twice the share of its aggregatelabor force employed in scientific re-search and development compared withthe country and Texas. Last, specializationin industries related to tourist activity—general merchandise stores, restaurantsand traveler accommodation services—isat least one and a half times higher in themetro than in the state and nation. Theonly vivid difference is the concentrationof workers in wired telecom services; themetro’s share of these workers exceeds the

national share by 112 percent and thestate share by merely 38 percent. Joblosses during the recent downturn weremitigated in San Antonio because of themetro’s low concentration of high-techindustries. Thus, the concentration of themetro’s base industries has held steady.

Fort Worth –Arlington. Fort Worth–Arlington is a major air and rail trans-portation hub in Texas with historic ties tooil, aircraft and aerospace product manu-facturers. Fort Worth’s employment sharesin aircraft, aerospace products and partsmanufacturing, communications equip-ment manufacturing as well as air and railtransportation are two to four times higherthan the state’s shares. This agglomerationbecomes more prominent when com-pared with the nation (Chart 6 ). Fort

Chart 5San Antonio Compared with the United States

National securityand internationalaffairs

Locationquotients

0

1

2

3

4

5

6

Footwear mfg.

Petroleumrefining

Generalmerchandisestores

Electronicshopping andmail-order houses

Insurancecarriers andrelated activities

Textile productmills

Not specifiedutilities

Electric, gas, and other utilities


Wired telecomcarriers

Savings institutionsincluding credit unions

Traveleraccommodation

Restaurants and otherfood services

Businesssupportservices

ScientificR&Dservices

Homehealth careservices


Chart 6Fort Worth Compared with the United States

Locationquotients


0

1

2

3

4

5

6

7

Support activitiesfor mining Electronic components

and mfg.

Machinery mfg. Wiredtelecommcarriers

Travel arrangementsand reservationsservices

Computer systemsdesign and relatedservices

Warehousing and storage

Communications, audioand video equipment mfg.

Aircraft and parts mfg.

Aerospaceproducts andparts mfg.

Railroad rolling stock

Air transportation

Railtransportation

Paperboard boxesand containers mfg.

Oilfield machinery mfg.


Exportable goods and

services are important

because they generate

out-of-state revenue and

stimulate state growth.


Michael E. Porter, Economic Development Quar-terly, vol. 14, February 2000, pp. 15–34.

3 Although the use of location quotients is preva-lent, this measure of the economic base canhave shortcomings. Some of the pitfalls ofregional analysis using location quotients areunderestimating the degree of geographic con-centration if the reference region is a netexporter of the good or service and overestimat-ing the degree of geographic concentration if thereference region is a net importer of the good orservice.

4 Integrated Public Use Microdata Series: Version3.0, by Steven Ruggles, Matthew Sobek, TrentAlexander, Catherine A. Fitch, Ronald Goeken,Patricia Kelly Hall, Miriam King and Chad Ron-nander, Minneapolis, Minn.: Minnesota Popula-tion Center, 2004. For more information, seewww.ipums.org.

5 The economic base is computed by adding sur-plus service-export employment (individualsemployed in producing services in excess oflocal demand) to total manufacturing and min-ing employment. Therefore, the share of Texasemployment included in the export base is

total manufacturing employment + total mining employment + {sum of [(LQ– 1/LQ) ×employment] for all service-providing sectors

with LQ above 1.1}

total Texas employment.6 For Texas and all its major metropolitan areas,

industries with high location quotients, whichhave been referenced in the text as key contrib-utors of their respective economic bases, werealso the largest employers in the state and itsmajor metros, unless otherwise noted.

7 Evidence from the 2000 and 2004 Bureau ofLabor Statistics data shows that the recessiondid not change the ordering of Texas’ basicindustries. The location quotients of severalhigh-tech industries declined slightly, but theirshares are still higher than U.S. shares.

8 For an in-depth analysis of the attributes of theTexas major metros and how each of them grewduring the 1990s, see, “Economic RecoveryUnder Way in Major Texas Metros,” by D’AnnPetersen and Priscilla Caputo, Federal ReserveBank of Dallas Southwest Economy, March/April 2004.

Worth–Arlington’s employment shares inthese same industries is more than threeto six times those of the nation. Since therecent recession, the composition of FortWorth’s leading industries has remainedunchanged.

El Paso. El Paso specializes in manu-facturing, trade and transportationbecause of its close ties with Mexico(Chart 7). More recently, the metro hasseen substantial growth in its high-wagemanufacturing and service sector. Gener-ally, the trade sector displays limited spe-cialization. As a result of El Paso’s locationalong the U.S.–Mexico border, however,the metro’s employment share in ware-housing and storage is twice that of Texasas well as the nation. The metro’s employ-ment shares in footwear, cut and sewapparel, textile and household appliancemanufacturing are more than 10 timeshigher than the state’s employmentshares. These shares are also high whencompared with the nation. Cut and sewapparel, footwear and household appli-ance manufacturing exceed nationalshares by eight times. Despite the highshares, the passage of NAFTA has led tomuch of this manufacturing going acrossthe border. Thus, the number employed inthese industries makes up only 4 percentof El Paso’s total employment. The largestemployers in El Paso today are still closelytied to the maquiladora industry acrossthe border but are a different set of indus-tries, including plastic products manufac-turers, electronic component and productmanufacturers, department stores, truck-ing, warehousing and storage firms.

ConclusionThe Texas economy thrives on a

diverse mix of industries. Once known asthe land of oil, cotton and cattle, Texas hasdeveloped into a high-tech hub. High-tech and energy sectors are the state’sdensest clusters, but Texas has manyother industries whose shares in the stateare higher than their shares in the nationand thus contribute to the state’s eco-nomic base. Nearly 35 percent of Texas’employment is in industries that can beclassified as basic, or exportable. In thesebasic industries, slightly less than half theworkers are engaged in producing goodsand services that satisfy nonlocaldemand. Exportable goods and servicesare important because they generate out-of-state revenue and stimulate stategrowth. Moreover, because clustersimprove efficiency and innovation, theformation and growth of clusters areimportant for Texas to maintain its com-petitive edge in this era of globalization.

Assanie is an assistant economist andYücel is a senior economist and vice presi-dent in the Research Department of theFederal Reserve Bank of Dallas.

Notes1 See “Competitive Advantage, Agglomeration

Economies and Regional Policy,” by Michael E.Porter, International Regional Science Review,vol. 19, no. 1 & 2, 1996, pp. 85–94.

2 For more information on what clusters are andhow they affect competition and innovation, see“Location, Competition and Economic Develop-ment: Local Clusters in a Global Economy,” by

Chart 7El Paso Compared with the United States

Locationquotients

Natural gasdistribution Services incidental

to transportation

Electroniccomponentsand products mfg.

Agricultural implements mfg.

Radio,TV and cablebroadcastingservices

Savingsinstitutionsincludingcredit unions

Nonferrous metalproduction and processing

Footwear mfg.

Householdappliance mfg.Cut and sew apparel mfg.

Warehousingand storage

Leather tanningand products mfg.

Plasticproducts mfg.

0

2

4

6

8

10

12

Departmentstores

Trucktransportation

Nationalsecurity andinternationalaffairs

Justice and publicorder activities

Restaurantsand other foodservices



SSince 1969, the Texas economy hasgrown rapidly, consistently matching orexceeding the growth of the nationaleconomy from one decade to the next.Real personal income growth rates inTexas matched the U.S. rates even duringthe oil bust years of 1979–89 andexceeded U.S. rates in 1969–79 and1989–2001 (Table 1). Measured by totalpopulation, growth in Texas was substan-tially greater in all periods.

The state’s largest metropolitanareas—Dallas–Fort Worth, Houston,Austin and San Antonio, which togethermake up what is known as the Texas Tri-angle—have contributed the largest partof this growth, especially since 1979. Out-side the Texas Triangle cities, real incomegrowth has failed to match U.S. growthsince 1979, although population hasexpanded somewhat faster.

This growth has improved Texas’ eco-nomic position relative to the rest of theUnited States. Texas moved from thenation’s fourth most populous state in1969 to second in 2001, trailing Californiabut ahead of New York and Florida. Interms of personal income, Texas hasmoved from the sixth largest state econ-omy in 1969 to the third largest today,behind California and New York.

The state’s large metropolitan areashave similarly moved up the ranking ofthe nation’s largest cities.1 Dallas–FortWorth, Houston and San Antonio mademost of their climb through these rank-

Economic Progress in the Texas Economy

Robert W. Gilmer

Table 1Growth of Population and Personal Income in Texas and the United States(Average percent per year)

Population

1969–1979 1979–1989 1989–2001

United States 1.1 1.0 1.2Texas 2.3 1.9 2.0Dallas–Fort Worth 2.4 3.0 2.6Houston 3.4 1.9 2.3Austin 4.1 3.8 3.9San Antonio 1.9 2.1 1.8Texas Triangle 2.8 2.5 2.5Rest of Texas 1.7 1.2 1.3

Personal Income

1969–1979 1979–1989 1989–2001

United States 3.7 3.0 2.9Texas 6.0 3.0 4.3Dallas–Fort Worth 5.4 4.7 4.7Houston 8.0 2.4 5.1Austin 2.3 5.6 8.3San Antonio 4.7 4.2 4.0Texas Triangle 6.5 3.8 5.0Rest of Texas 5.2 1.7 2.9

NOTE: Based on 1999 metropolitan area definitions of the Office of Management and Budget. Dallas–Fort Worthand Houston use the consolidated metro area definition.

SOURCES: Bureau of Economic Analysis; author’s calculations.

In terms of personal

income, Texas has

moved from the sixth

largest state economy

in 1969 to the third

largest today, behind

California and New

York.


ings between 1969 and 1979 (Table 2).2

Since 1979, Dallas–Fort Worth and Hous-ton have shared the eighth through tenthspots in population and personal income,while San Antonio moved slowly upwardto 32nd in population and 35th in per-sonal income.

Austin, however, made steady anddramatic gains. In 1969, at No. 75 in pop-ulation, Austin was the size of Canton, Ohio, or Fort Wayne, Ind. But by2001, at 39th, Austin’s population com-pared favorably with that of Nashville orNew Orleans. During the same period,Austin surged from 86th to 37th in per-sonal income.

Table 3 summarizes the contributionof these different metro areas to Texas’personal income growth. Except for the oil bust years, Houston contributed nearly 30 percent of growth, and Dallas–Fort Worth’s growth exceeded Houston’sby the late 1970s. San Antonio’s growthcontribution held steady at 6 to 8 percent,while Austin’s doubled from 4.1 percent to

8.3 percent. The combined metro areas,collectively designated the Texas Trianglein the table, accounted for three-fourthsof the state’s income growth between 1989and 2001.

In this article, we will measure the

success of the Texas economy not by itssize, growth rates or ranking, but by thestate’s ability to improve the welfare of itscitizens. In particular, we will look at thestate’s ability to raise its per capita incomelevels to those of the nation—to join andperhaps outperform the nation’s main-stream. Income per person presents anumber of flaws as a measure of generalwelfare, but it serves here as a widely rec-ognized and useful summary of the stan-dard of living.3

Texas Per Capita Income In 1969, per capita income in Texas

was $3,373, or 87.7 percent of the U.S.level. Fueled by the oil boom after 1973,Texas’ per capita income grew rapidly tobriefly exceed that of the United States by1981–82 (Chart 1). The 1980s oil, bankingand real estate bust quickly erased thesegains, and by the end of the decade, stateper capita income had returned to 87.9percent of the U.S. level.

The 1990s brought new advances rel-ative to the nation as oil, high tech and afree trade- and maquiladora-inspiredboom along the Texas –Mexico borderproduced another burst of Texas eco-nomic growth. By 1998, Texas per capitaincome returned to 94.4 percent of U.S.levels and made no further progressthrough 2001.

We can examine Texas per capitaincome growth both geographically andby the components of income—wagesand salaries, proprietor’s income, prop-erty income, transfers and other sources.By component, the most interesting

Table 2Rank of Texas Triangle Metro Areas in United States by Population and Personal Income

Population

1969 1979 1989 2001

Dallas–Fort Worth 12 9 9 8Houston 13 10 10 9Austin 75 63 63 39San Antonio 37 33 33 32Texas Triangle 4 4 4 3

Personal Income

1969 1979 1989 2001

Dallas–Fort Worth 13 10 9 8Houston 16 9 10 9Austin 86 69 55 37San Antonio 45 39 38 35Texas Triangle 9 4 3 3


Table 3Contribution to Texas Personal Income Growth

Percent

1969–1979 1979–1989 1989–1999 1989–2000 1989–2001

Texas 100.0 100.0 100.0 100.0 100.0Dallas–Fort Worth 23.6 31.9 30.7 31.3 30.6Houston 28.5 23.8 28.2 28.6 29.7Austin 4.1 6.0 8.7 8.6 8.3San Antonio 6.3 8.1 7.1 6.9 7.0Texas Triangle 62.6 69.8 74.6 75.4 75.5Rest of Texas 37.4 30.2 25.4 24.6 24.5


Chart 1Convergence of Texas to U.S. Per Capita Income Levels

Index: U.S. per capita income = 100

50

120

110

100

90

80

70

60

’69 ’73 ’77 ’81 ’85 ’89 ’93 ’97 ’01

Rest of Texas

Triangle cities

Texas

SOURCE: Bureau of Economic Analysis.


Chart 2 shows the path of the fourcities since 1969 in terms of incomegrowth relative to the nation’s. The gainsand losses of the boom and bust in oil andreal estate are visible in all four cities, butmost notably in Houston and Austin. Allcities made gains in the 1990s, especiallyAustin. San Antonio made the leastprogress, despite beginning from the low-est per capita base. The two high-techmetros began losing ground in relation tothe United States well before the nationalrecession began in 2001, with Austinpeaking at 110 percent of U.S. levels in1999 and Dallas–Fort Worth at 112 per-cent in 2000. Houston reached 115 per-cent of U.S. per capita income in 2001.San Antonio stood at 88 percent.

The fact that the four cities have suchdifferent income levels and very differentbehavior over time might seem surprisingin light of their geographic proximity. But,

in fact, it may be this very proximity thatguarantees their different personalities.Because no pair of cities in the Texas Tri-angle is more than 240 miles apart, eachhas assumed a role in the state economythat sets it apart and makes it distinctfrom the others.6

Dallas–Fort Worth. Dallas–Fort Worthis a major inland transportation hub anddistribution center for Texas, Louisiana,Arkansas and Oklahoma and claims theworld’s fifth busiest airport. Following theoil bust, Dallas emerged as the state’sbanking and financial center. Dallas andFort Worth also have a significant presenceof oil-related activity, notable on any standard except that set by Houston.High-technology industries, especiallytelecommunications, became a majorcenter of growth in the 1990s.

Houston. Houston’s bread and butterremains oil and natural gas, with oil pro-

results come from the growth of wagesand salaries and proprietor’s income. Thegeographic designation focuses largely onthe Texas Triangle cities, which havefueled both the state’s growth and most ofits recent convergence to U.S. per capitaincome levels.

Framework for Analysis The general framework used here is

shown in Table 4, which summarizes percapita income growth in Texas by compo-nent of income, geographic area and timeperiod from 1969 to 2001.4 The data arepresented as percentage point contribu-tions to average annual real per capitaincome growth in each region and timeperiod.5

For example, the growth of per capitaincome in Texas from 1969 to 1979 aver-aged 3.6 percent per year, with most of thegrowth (3 percent per year) coming fromwages and salaries per capita and smallercontributions from property income (0.2percent), transfer payments (0.2) andother per capita income (0.4). Proprietor’sincome per capita grew more slowly thanother components, reducing the growthrate by 0.2 percent.

The components of income defini-tions follow standard conventions foraccounting for personal income in thenational income and product accounts.The definitions are fairly obvious: non-farm wages and salaries; farm and non-farm proprietor’s income earned by soleproprietorships, partnerships and tax-exempt corporations; property incomefrom dividends, rent and interest; andtransfer payments for no current servicesrendered. The “other income” category isa residual made up mainly of benefitspaid to wage and salary workers, but italso includes a residence adjustment forworkers who live and work in different areas.

The rationale for the geographic focuson the Texas Triangle has partly been dis-cussed above, primarily because three-fourths of the region’s personal incomegrowth came from these metro areas after1989. Also, most of the forces drivingincome convergence have come from theTriangle cities. While per capita incomelevels were, on average, well abovenational norms and rising through the1990s within the Triangle, they werefalling back to near 70 percent outside of it.

Table 4Growth Rate of Real Per Capita Personal Income and Factors Contributing to Its Growth(Average percent per year)

Component Percentage Point Contribution Per Capita

Personal Nonfarm wages Proprietor’s Property Transfer Otherincome and salaries income income payments income

1969–1979United States 2.6 1.6 –.1 .3 .4 .4Texas 3.6 3.0 –.2 .2 .2 .4Dallas–Fort Worth 2.9 2.1 .1 .2 .2 .4Houston 4.4 4.2 –.1 –.1 .1 .3Austin 3.4 2.7 –.1 .3 .1 .5San Antonio 2.7 1.3 .1 .3 .4 .6El Paso 1.4 .9 .2 .3 .6 –.6Texas Triangle 3.5 2.9 0 .1 .1 .4Rest of Texas 3.4 2.8 –.5 .4 .3 .4

1979–1989United States 2.0 1.4 0 .7 .1 –.1Texas 1.1 .2 0 .7 .2 0Dallas–Fort Worth 1.6 1.1 .1 .4 0 –.1Houston .5 –.8 .5 .6 .2 0Austin 1.9 1.7 –.6 .6 0 .1San Antonio 2.0 1.1 .1 .7 .1 .1El Paso 1.7 .2 –.2 .7 .1 .9Texas Triangle 1.2 .3 .2 .6 .1 0Rest of Texas .6 –.6 –.5 1.1 .4 .2

1989–2001United States 1.7 1.8 .1 –.1 .2 –.2Texas 2.2 2.4 .4 –.4 .1 –.3Dallas–Fort Worth 2.0 2.4 .2 –.3 .1 –.3Houston 2.7 2.3 1.0 –.6 .1 –.2Austin 2.9 4.2 .1 –.7 –.1 –.6San Antonio 2.2 2.1 .7 –.3 .2 –.5El Paso 1.7 .9 .8 –.2 .5 –.3Texas Triangle 2.4 2.5 .5 –.4 .1 –.3Rest of Texas 1.6 1.5 0 –.4 .4 0



ducers, oil services and machinery com-panies, refineries and petrochemicalsdirectly or indirectly accounting for halfthe metro area’s jobs. The Texas MedicalCenter and Johnson Space Center, alongwith companies such as Continental Air-lines, American General Insurance andHP/Compaq, help define the non-oil partof Houston’s economy. Houston is thestate’s major deepwater port—the secondlargest in the country based on tonnage—and home to the state’s international busi-ness community.

Austin. Because it is the state capitaland site of the University of Texas’ maincampus, Austin’s major strength has his-torically been a robust government sector.Beginning in the late 1960s, Austin begandeveloping a significant presence in hightechnology: IBM in 1967, Texas Instru-ments in 1969 and Motorola in 1974. Thearrival of chipmaker-consortium Semat-ech in 1988 provided the momentum forthe 1990s. Today, about 120,000 employ-ees—25 to 30 percent of the local work-force—are tied to technology industries,and Dell Inc. has emerged as the city’smost important technology employer.Austin is also renowned for its musicindustry. Billed as the “Live Music Capitalof the World,” the city sponsors a numberof festivals and conventions based onmusic.

San Antonio. San Antonio’s historicrole has been as the distribution point forSouth Texas and northern Mexico, a rolethat has grown with the rapid expansionof the maquiladora industry and theimplementation of the North AmericanFree Trade Agreement. Tourism is a major

industry, with such features as FiestaTexas, SeaWorld, the River Walk, El Mer-cado and others. Lackland Air Force Base,Fort Sam Houston and Randolph AirForce Base provide a major military pres-ence.

One could speculate that if Texas’geography had been only slightly differ-ent—with navigable rivers or a saltwaterinlet that cut into the heart of the state—the four cities could easily have been one.The port, the inland distribution pointand the political capital would all havebeen colocated. Because the four Trianglecities play such different economic roles,adding up their current populations pro-duces a not far-fetched approximation ofwhat might have been a single metro area.The combined ranking of the Trianglecities (bottom of Table 2) shows that sucha combination would rank third among allU.S. consolidated metro areas—behindNew York and Los Angeles but ahead ofChicago—in both personal income andpopulation in the 1990s.

It is difficult to generalize about thearea outside the Triangle, or to easily char-acterize an area that includes cities as dif-ferent as El Paso, Amarillo, Texarkana andBeaumont. The decline of agriculturethroughout the second half of the 20thcentury played a large role in the region’spoor performance.

In addition, the Texas–Mexico borderacts as a drag on any measure of eco-nomic progress or welfare in the state,including per capita income. Gilmer,Gurch and Wang have already examinedthe Texas border cities using the sameframework employed here.7 The bordercities’ average per capita income is only50 to 60 percent of the national averageand has only occasionally matched orexceeded the state’s overall growth rate(such as Laredo in the 1990s). El Paso, byfar the largest Texas–Mexico border city,saw its per capita income fall from 73 per-cent of the U.S. average in 1969 to 63 per-cent in 2001. Although the border sawgains in income and jobs in the 1990s,rapid population growth due to highbirthrates and in-migration meant livingstandards did not improve nearly as muchas overall growth statistics might indicate.

How Income Grew in TexasExcept for the oil bust years, Texas’

per capita income outgrew the nation’s by

The fact that the four

Texas Triangle cities

have such different

income levels and

very different behavior

over time might

seem surprising

in light of their

geographic proximity.

Chart 2Convergence of Texas Triangle Cities to U.S. Per Capita Income Levels

Index: U.S. per capita income = 100

’69 ’73 ’77 ’81 ’85 ’89 ’93 ’97 ’01

San Antonio

Austin

Dallas–Fort Worth

Houston

70

80

90

100

110

120

130

140

SOURCE: Bureau of Economic Analysis.


a significant margin (see Table 4). The dif-ference was a full percentage point from1969 to 1979 (3.6 versus 2.6) and by half apercentage point from 1989 to 2001 (2.2versus 1.7). With the oil bust and recoveryfactored in, however, the difference infavor of Texas narrows to 0.2 percent (2.3versus 2.1 over the 32-year period), andper capita income rises from 88 percent to94 percent of the national average.

Also except for the oil bust years,most of the growth in Texas’ real percapita income came from increases in realwages and salaries per capita—83 percentfrom 1969 to 1979 and 109 percent from1989 to 2001. Only during the years of theoil and banking crisis did real wages andsalaries fail to contribute strongly toincome growth; only 17 percent of growthcame from that source from 1979 to 1989.Growth in property income (most proba-bly in the first half of the 1980s) was themajor factor contributing to incomegrowth during the decade of the down-turn.

Proprietor’s income makes its largestcontribution from 1989 to 2001. Houstonhas the strongest contribution from theself-employed in this period (1 percent)and during the previous period as well(0.5 percent). In 16 cities in Texas andLouisiana, all with strong ties to oil, thefirst result of the oil bust was a large num-ber of new “proprietors,” presumably newbusinesses started by people unemployedby the downturn.8 This forced entrepre-neurship was followed in the late 1980sand early 1990s by rapidly growing propri-etor’s income, the fruit of the businessesthat succeeded. The often-used analogy ofa forest fire leaving behind the seeds forthe forest’s regeneration seems to apply toTexas in recent years, with entrepreneur-ship sowing the seeds. On average, propri-etor’s income contributed 0.5 percent toper capita income growth in Texas Trian-gle cities in the 1990s.

Property income (dividends, rent andinterest) was the biggest contributor toper capita income growth during the oilbust and recovery years. The 1980s saw alarge run-up in property values, which fellback slowly late in the decade but droveup rental values, and a sharp hike in inter-est rates due to inflation and tight mone-tary policy increased income from inter-est-earning sources. The contribution ofproperty income is small from 1969 to

1979 and negative from 1989 to 2001. Other income per capita makes its

largest contribution from 1969 to 1979, isneglible from 1979 to 1989 and turnsslightly negative in the most recentperiod.

A Closer Look at Wageand Salary Growth

Because wages and salary growth percapita account for such a large share ofTexas per capita income, we will examineit more closely. We can divide wages andsalaries per capita (WS/P) into two parts:wages and salaries per employee (WS/E)and the employment population ratio(E/P).

WS/P = WS/E × E/P

Further, we can offer two reasons for thegrowth of wages and salaries peremployee: (1) improvements in the indus-try mix that allow more workers to moveinto higher-paying industries, or (2) spe-cific advantages the region offers in

resources, labor supply, infrastructure orother local factors. This region-specificadvantage is called differential regionalearnings.9

WS/P = WS/E × E/P = industry mix ×differential regional earnings × E/P

Table 5 summarizes the contributionof each of these elements to real per capitaincome.10 The first column is wages andsalaries per worker; the second and thirdcolumns divide this category into twoparts. The fourth column is the employ-ment population ratio, or jobs per capita.

Industry mix was a significant factorin all areas and in every period. Texas wasclearly shedding low-wage jobs andreplacing them with better-paying jobsthroughout the entire period.

We also see gains from differentialregional earnings in the two periods ofrapid growth. In the 1990s the Texas Triangle cities added 0.6 percent per yearto per capita income thanks to theseadvantages. The measure highlights the

Table 5Impact on Per Capita Income of Industry Mix, Differential Regional Earnings and Jobs Per Capita

Percentage Point Contribution to Annual Growth Rate

Wages and salaries Industry Differential Jobs perper worker mix regional earnings capita

1969–1979Texas 1.5 1.3 .2 1.5Dallas–Fort Worth .8 1.1 –.2 1.3Houston 1.9 1.1 .8 2.3Austin 1.2 1.2 0 1.4San Antonio 1.2 1.5 –.3 .2Texas Triangle 1.4 1.2 .2 1.5Rest of Texas 1.5 1.5 0 1.2

1979–1989Texas .3 .8 –.5 –.1Dallas–Fort Worth .9 .9 .1 .2Houston –.2 .7 –.8 –.6Austin 1.2 1.2 0 .5San Antonio .5 .8 –.3 .6Texas Triangle .4 .8 –.4 –.1Rest of Texas –.3 .8 –1.0 –.4

1989–2000*Texas 1.8 1.4 .4 .8Dallas–Fort Worth 2.1 1.3 .8 .7Houston 1.9 1.5 .3 .5Austin 3.7 1.3 2.4 1.5San Antonio 1.2 1.4 –.2 1.1Texas Triangle 2.1 1.4 .6 .8Rest of Texas .7 1.0 –.3 .8

* Data extend only to 2000 due to a change in the distribution of jobs from the Standard Industrial Classification to North American Industry Classification System in 2000, making it impossible to compare 1989 with 2001.

NOTE: Differences due to rounding error.



state’s booms and busts: Houston added0.8 percent per year from 1969 to 1979,which turned to –0.8 percent the follow-ing decade. Large regional differentials inAustin (2.4 percent) and Dallas–FortWorth (0.8 percent) mark the 1990s techboom. A look back at Chart 2 shows thatthese cities were already giving back someof their tech gains by 2001.

During the two decades of stronggrowth, the state generated jobs fasterthan the rate of population growth,despite rapid in-migration (Table 6). Percapita job growth has occurred inside andoutside the Triangle cities despite the fact, as mentioned above, that the bordercities were unable to attain job growthmuch faster than population growth. Thiscontributed 1.5 percent per year to Texasper capita income growth (as seen in column 4 of Table 5) from 1969 to 1979and 0.8 percent from 1989 to 2000. Theslight decline in the 1980s (–0.1 percent)was primarily due to slower job growth in Houston and areas outside the Triangle.

Summary and Conclusions Measured by standards of population,

employment and income growth, theTexas economy has outperformed theU.S. economy since 1969. As shown inTable 7, by 2001 the state as a whole hadraised its per capita income to 94 percentof the national average, up from 88 per-cent in 1969. Over the same period, theaverage annual growth rate of per capitaincome was 2.3 percent for Texas versus2.1 percent for the United States.

Economic progress has been unevenover time. The oil boom briefly pushedTexas per capita income above thenation’s in 1981–82. In the subsequentcollapse of oil, banking and real estate,Texas fell back to almost its 1969 positionrelative to the United States. Most subse-quent progress has come since 1989, andit primarily can be attributed to more jobsavailable to the general population and an improving mix of jobs with highersalaries.

Table 7 also indicates the uneven geographic progress. In fact, the forces ofconvergence to U.S. levels have mostlycome from the Texas Triangle metropoli-tan areas of Dallas–Fort Worth, Houston,Austin and San Antonio. All these citieshave outperformed the United Statessince 1969, with the most dramatic gains

Table 6Employment and Population Growth, 1969–2001

Job Growth*

1969–1979 1979–1989 1989–2001

United States 2.2 1.8 1.5Texas 3.8 1.8 2.7Dallas–Fort Worth 3.6 3.2 3.0Houston 5.8 1.3 2.7Austin 5.5 4.4 4.9San Antonio 2.0 2.6 2.7Texas Triangle 4.3 2.5 3.0Rest of Texas 3.0 .8 2.0

Population Growth*

1969–1979 1979–1989 1989–2001

United States 1.1 .9 1.2Texas 2.3 1.9 2.0Dallas–Fort Worth 2.3 3.0 2.6Houston 3.4 1.9 2.3Austin 4.1 3.8 3.9San Antonio 1.9 2.1 1.8Texas Triangle 2.8 2.5 2.5Rest of Texas 1.7 1.2 1.3

Jobs Per Capita*

1969–1979 1979–1989 1989–2001

United States 1.1 .8 .3Texas 1.5 –.1 .7Dallas–Fort Worth 1.3 .2 .4Houston 2.3 –.6 .4Austin 1.4 .5 1.0San Antonio .2 .6 .9Texas Triangle 1.5 –.1 .5Rest of Texas 1.2 –.4 .7

* Annualized growth rates.


During the two decades of strong growth,

Texas generated jobs faster than the rate

of population growth, despite rapid

in-migration.


growth and the locational advantages of thestate’s largest metro areas.

4 The framework was developed by Daniel H. Gar-nick. See “Accounting for Regional Differencesin Per Capita Personal Income Growth,1929–79,” by Daniel H. Garnick, Survey of Cur-rent Business, vol. 62, September 1982, pp.24–34, and “Accounting for Regional Differ-ences in Per Capita Income Growth: An Updateand an Extension,” by Daniel H. Garnick andHoward L. Friedenberg, Survey of Current Busi-ness, vol. 70, January 1990, pp. 29–40.

5 Constant dollars are obtained by deflating withthe personal consumption expenditure deflator(1996 = 100) for all areas.

6 “The Simple Economics of the Texas Triangle”(January 2004) and “The Texas Triangle asMegalopolis” (April 2004), both by Robert W.Gilmer, in Houston Business, Federal ReserveBank of Dallas.

7 “Texas Border Cities: An Income Growth Per-spective,” by Robert W. Gilmer, Matthew Gurchand Thomas Wang, The Border Economy, Fed-eral Reserve Bank of Dallas, June 2001, pp. 2–5.

8 “Finding New Ways to Grow: Recovery in the OilPatch,” by Robert W. Gilmer, Houston Business,Federal Reserve Bank of Dallas, July 1996.

9 The actual calculation of industry mix and differ-ential regional earnings is spelled out carefully inGarnick and Friedenberg (1990). The calculationdepends on the definition of hypothetical income(H), total wages and salaries that would havebeen earned in Texas if compensation were paidat the national rate in each industry. Hypotheti-cal income was calculated using the wage andsalary employment categories in the Bureau ofEconomic Analysis’s Regional Economic Infor-mation System, essentially a one-digit definitionin the Standard Industrial Classification. Usingthis definition,

WS/P = industry mix × differential regional earnings × E/P = H/E × WS/H × E/P.

10 The data in Table 5 extend only to 2000 becauseof the change in the industrial classification sys-tem from the Standard Industrial Classification

to the North American Industry ClassificationSystem, beginning in 2001. This made it impos-sible to compare the distribution of jobs andincome by industry in 1989 and 2001.

coming out of Austin. The addition of a large high-technology workforce to a stable, if less-well-paid, government anduniversity base fueled both rapid growthand rising per capita income in the statecapital. Except for San Antonio, all thecities enjoy living standards above theU.S. average.

The uneven nature of Texas’ eco-nomic history makes it difficult to predictfuture progress. The geographic concen-tration of growth seems unlikely tochange, but the state’s advantages relativeto the rest of the nation (as measured by differential regional earnings) weredominated by the oil boom from 1969 to 1979 and to some extent by the high-tech expansion of 1989–2001. Advantageswere concentrated first in Houston, then in Austin and Dallas–Fort Worth.Predicting the source or location of thenext great round of expansion is impossi-ble.

However, since 1969 Texas’ costadvantages, tax advantages, climate andlifestyle have prepared the ground for fur-ther growth and development, includingperiodic excesses. These Sunbelt advan-tages should persist, making renewedeconomic expansion in Texas and contin-ued progress in raising the state’s livingstandards simply a matter of time.

Gilmer is a vice president at the FederalReserve Bank of Dallas.

Notes1 The statistics for Dallas–Fort Worth and Hous-

ton use their consolidated metropolitan statisti-cal area definition throughout this article. Theranking of metro areas includes consolidatedmetropolitan statistical areas (CMSAs) but thenexcludes all the parts of these CMSAs (metro-politan and primary metropolitan statisticalareas) in the subsequent ranking process.

2 The end years used here—1969,1979,1989 and2001— are all peak years in the U.S. businesscycle. Although Texas and its metro areas didnot always follow the U.S. cycle, particularly inthe 1980s, these years were typically times ofeconomic expansion for Texas, making compar-isons to the U.S. economy appropriate.

3 The most notable flaw in the use of per capitaincome as a measure of welfare is that it tells usnothing about the size distribution of incomeamong the population. However, this articledivides per capita income into enough cate-gories by component and geography to givesome insight into how income growth is affectedby regional wage levels, job growth, population

Table 7Performance of Regions of the Texas Economy

2001 per Annual growth ratecapita income Percent of 1969–2001

(dollars) U.S. level (percent per year)

United States 30,413 100 2.1Texas 28,472 94 2.3Dallas–Fort Worth 33,247 109 2.2Houston 34,916 115 2.5Austin 31,511 104 2.8San Antonio 26,887 88 2.3Texas Triangle 32,897 108 2.4Rest of Texas 21,357 70 1.8



OOver the past 10 years, trade betweenthe United States and Mexico hasboomed, partly because of the significantreduction in tariffs from NAFTA and thestrong growth in the maquiladora indus-try. Along with the expansion in trade,there has been strong population growthalong the northern border of Mexico.Generally, the population in Mexican bor-der cities is significantly larger than in thecorresponding U.S. sister cities. Moreover,the South Texas border metros are a shortdrive from the industrial city of Monter-rey, which had a population of 3.8 millionin 2000. The large and growing populationon the Mexican side of the border repre-sents an important consumer base forretail stores in U.S. border towns.

While commercial trade between theUnited States and Mexico is well docu-mented, less is known about the size ofthe nations’ cross-border retail trade.Though small in comparison with com-mercial trade, this retail trade is a signifi-cant part of many border city economies.In 2003 alone, there were more than 38million noncommercial crossings at thebridges along the Texas–Mexico border.Many of these individuals were coming topurchase goods to take back to their homecountry. Due to differences in nationalpolicies such as environmental laws, taxesand consumer safety regulations, peoplecross daily to purchase goods and serviceson both sides of the border.

Since most of the retail trade con-ducted on the U.S. side of the border isdone in cash, it is difficult to document theshare of retail spending by Mexicannationals. In this article, we use a simpleconsumption function to estimate theamount of retail spending that is essen-tially exported to Mexico via cross-bordershoppers.1 Since the true amount spent byMexican nationals is not known, it is diffi-cult to estimate the accuracy of our meas-ures. Theory tells us, however, that metro

areas having the biggest share of theirretail sales going to Mexican nationals willbe impacted the most by large swings inthe value of the peso. We thus check thatour estimates are consistent with theeffects on local retail sales of movementsin the real dollar/peso exchange rate.

Previous Research on Border RetailTraditionally, the border has been a

region of fast population and job growthcompared with the rest of the United Statesand Mexico. The Border IndustrializationProgram—enacted in 1965 by the Mexicangovernment after the United States endedthe Bracero Program—gave birth to themaquiladora industry, which in turn inten-sified the border region’s growth, not only

in Mexico but also on the U.S. side due toincreasing border interlinkages. Themaquiladora industry has been the maineconomic growth driver along theTexas–Mexico border.

Several studies have addressed theissue of cross-border retail trade as part ofa larger question of the maquiladora in-dustry’s impact on the regional economiesof U.S. border cities. The first studies on thesubject date back to the early 1970s andindicate that a significant portion ofmaquiladora salaries was spent on the U.S.side of the border, mainly on food andclothing. More specifically, one study esti-mates that a 10 percent increase inmaquiladora employment translates into a23 percent increase in retail sales inBrownsville, a 13 percent increase inLaredo, an 11 percent increase in El Pasoand a 7 percent increase in McAllen.2

Perhaps the first researcher to studythe impact of the maquiladoras along theTexas border in a comprehensive mannerwas J. Michael Patrick.3 His main conclu-sion regarding cross-border retail tradeactivity is that growth in the maquiladoraindustry in Mexico stimulates U.S. borderjob growth mostly in the retail and servicesectors, not in the manufacturing sectoras commonly perceived.

One of the first studies to quantify theimpact of Mexican nationals on retailtrade on the U.S. side of the border wasdone by the San Diego Chamber of Com-merce in 1979.4 Through surveys, thestudy estimated that 7.5 percent of SanDiego’s retail sales ($407 million) could beattributed to Mexican nationals. In 1993,according to a study by the San Diego Dia-logue, about 42 percent of the people whocrossed into San Diego were Mexicannationals with the main purpose of shop-ping. They accounted for $2.8 billion inretail sales.5

More recently, in 2002, Charney andPavlakovich-Kochi estimated the eco-

Texas Border Benefits fromRetail Sales to Mexican NationalsKeith R. Phillips and Roberto Coronado

El P

aso

Conv

entio

n &

Vis

itors

Bur

eau


nomic impact of Mexican visitors to theeconomy of Arizona. They found thatMexican visitors spent $962 million, withthe vast majority in department stores (41percent) and grocery stores (25 percent),mostly in border counties.6 Similarly, onthe Texas–Mexico border, the Center forBorder Economic Studies at the Univer-sity of Texas – Pan American estimatedthat total expenditures by Mexican visitorsin the lower Rio Grande Valley amountedto $1.4 billion in 2003.7

Other studies have focused on theimpact of exchange rate fluctuations onU.S. border retail sales. For instance, Diehlconcludes that the 1982 Mexican eco-nomic crisis that triggered peso devalua-tion stunned South Texas retailers by cut-ting retail sales as much as 80 to 90 per-cent in many border businesses.8 Simi-larly, Patrick and Renforth estimate,through the use of almost 4,000 surveys,that the 1994 peso devaluation resulted ina strong 41.8 percent decline in retailsales, but the results varied by city, storetype, distance from the border and rela-tive domestic market size.9 Gerber docu-ments the relationship between pesovalue fluctuations and total taxable salesin San Diego and Imperial counties,where he finds that an unanticipated 10percent decline in the value of the pesodepresses total taxable sales by approxi-mately 1 percent in San Diego County and2.22 percent in Imperial County.10

Many of the studies, however, areregion- and time-specific, making com-parisons across regions and over time dif-ficult. Also, many of the studies were doneusing time-consuming, labor-intensive,and thus expensive, survey techniquesthat would be difficult to perform consis-tently over time and across regions. Toovercome these limitations, we use a sim-ple consumption function approach thatproduces a consistent annual time seriesof exported retail sales for the four metro-politan statistical areas (MSAs) on theTexas–Mexico border.

Using a Different ApproachPhillips and Manzanares propose a

simple model in which it is assumed thatindividuals spend a fixed proportion oftheir income on consumption, or in thiscase, retail sales.11 For instance, they findthat from 1986 to 1998 retail sales as afraction of personal income in Texas aver-

aged 46 percent. For each of the four bor-der MSAs, they multiplied 0.46 by totalpersonal income to get an estimate ofretail sales purchased by the local popula-tion and then subtracted sales to localsfrom total sales to get net exported retailsales. If the value of net exported retailsales is negative, that means more localincome is spent outside the local econ-omy than income spent by outsiders inthe local community. While it is evidentthat many Mexican nationals cross theborder to shop, U.S. citizens also crossinto Mexico to dine at restaurants and tobuy local handicrafts, medicines, liquor,dental services and other products andservices. Border residents also vacationand shop at other destinations in theUnited States. Remittances to familymembers in Mexico can also reduce theamount of local income spent on localretail goods and thus reduce net exportedretail sales.

Using a constant fraction of local per-sonal income to estimate the amount thatlocals spend on retail—and using thisamount to estimate net exported retail—produces reasonable results. However, wecan further refine the model by decom-posing personal income into three com-ponents, allowing the coefficient on eachcomponent to differ. The border regionhas a low employment-to-populationratio due to its young labor force and highunemployment rates. It also has persist-ently low per capita personal income yetstrong job growth rates. If these factorsplay differing roles in retail spending, it isimportant to separate them out. We dividepersonal income (Y) as follows:

where POP is population and thus Y/POP isper capita income, POP/EMP is the inverseof the employment-to-population ratio andEMP is total employment. We then try toestimate the impact of the three compo-nents of personal income on retail salesacross the 23 non-border Texas MSAs. Weuse quarterly retail sales data at the metrolevel from 1978 to 2001, available from theTexas comptroller’s office. Annual personalincome for metro areas (less contributionsfor social insurance) from 1978 to 2001 isavailable from the Commerce Depart-ment’s Bureau of Economic Analysis.

Table 1Border Exported Retail Sales,1978–2001

Average share(percent)

Brownsville 25.7El Paso 11.3Laredo 51.1McAllen 35.6

SOURCE: Authors’ calculations.

We use the results from the model toestimate exported retail sales for the fourTexas border MSAs. Table 1 reports theaverage share of exported retail sales forthese four areas during our estimationperiod. According to our results, in 2001,Mexican shoppers accounted for morethan $2 billion in retail sales, representing0.75 percent of total retail sales in Texas. In2001, McAllen was the biggest net ex-porter of retail sales to Mexicans, withalmost $1 billion in sales, representing 33percent of its total local retail trade activ-ity. Laredo came in second with $540 mil-lion in exported retail sales, or 39 percentof total retail sales. Brownsville registered$256 million (16 percent of total retailsales), while El Paso, the biggest of thefour cities in terms of population, ex-ported only $215 million (6 percent) toMexican nationals. El Paso’s figure is wellbelow its average exported retail sales of11.3 percent and is primarily due to thecontracted maquiladora activity south ofthe border. Ciudad Juárez registered itsworst maquiladora performance in 2001and 2002, with employment decliningalmost 25 percent.

On average over the 1978–2001period, Mexican nationals accounted for1.6 percent of Texas retail sales, or $5.1million on a daily basis. Chart 1 showsthat over time Laredo has the highestshare of exported retail sales to actualtotal sales, followed by McAllen, Browns-ville and El Paso.

Sensitivity to Exchange Rate SwingsAlthough there is no straightforward

way to determine the accuracy of ourresults, retail sales from Mexican nation-als should be sensitive to swings in thevalue of the peso. These swings representprice shocks for Mexican nationals shop-ping on the U.S. side, and border retailersknow that sharp declines in the peso’s

Y YPOP

POPEMP EMP= × × ,,Y ( (

((

Chart 2Border Retail Sales Are Closely Related to Real Exchange Rate

Total retail sales index, 1978 = 100 Real exchange rate

Chart 1Exported Retail Sales Shares

Percent of total retail sales Real exchange rate

value result in a sharp drop in Mexicanshoppers. Under our model, exportedretail sales seem to be responsive tochanges in exchange rate (see Chart 1).

If exported retail sales represents asignificant portion of total retail sales,changes in the value of the peso shouldhave statistically significant impacts ontotal retail sales. To asses this, we performsome statistical tests on the sensitivity ofoverall retail sales to changes in the valueof the peso. Results show that, in all MSAsbut El Paso, changes in the real exchangerate have statistically significant impactson total local retail sales. The magnitudeof the impact was the largest in Laredo.Since our results show that El Paso had thesmallest share of its retail sales going toMexican nationals and Laredo had thelargest, these results are consistent withour previous findings (Chart 2).

OutlookIn mid-2005 the real value of the peso

was above its 20-year average and themaquiladora industry was continuing tobounce back from its downturn in2001–03. Both of these factors should con-tinue to stimulate growth along the Texasside of the border. Looking to 2006, Mex-ico is hoping to have its second consecu-tive presidential election without a pesodevaluation. The Texas border commu-nity is hoping for the same, as its economyebbs and flows with the movements in thevalue of the peso and the accompanyingwaves of Mexican shoppers.

Phillips is a senior economist and policyadvisor at the San Antonio Branch andCoronado is an assistant economist at theEl Paso Branch of the Federal Reserve Bankof Dallas.

Notes1 For a more detailed version of this article, please

see “Exported Retail Sales along the Texas–Mex-ico Border,” by Keith R. Phillips and RobertoCoronado, Federal Reserve Bank of Dallas work-ing paper (forthcoming).

2 See “Maquiladoras along the Texas–Mexico Bor-der: An Econometric Evaluation of Employmentand Retail Sales Effect on Four Texas BorderSMSAs,” by Richard J. Holden, Texas Depart-ment of Community Affairs, Regional EconomicDevelopment Division, February 1984.

3 See “The Economic Impact of Maquiladoras onBorder Development: A Rio Grande Valley CaseStudy— Some Preliminary Findings,” by J.Michael Patrick and Roland S. Arriola, paperpresented at the Western Social Science Associ-ation meeting, El Paso, Texas, 1987; “TheEmployment Impact of Maquiladoras along theU.S. Border,” by J. Michael Patrick, in TheMaquiladora Industry: Economic Solution orProblem?, ed. Khosrow Fatemi, New York:Praeger Publishers, 1990, pp. 31–35; and “TheImpact of NAFTA on Border Maquiladora andIndustrial Activity,” by J. Michael Patrick, Tech-nical Report, Center for Entrepreneurship andEconomic Development, University of Texas–Pan American, 1991.

4 “Mexican Impacts on Retail Sales,” San DiegoChamber of Commerce, San Diego EconomicBulletin, vol. 27, no. 6, 1979.

5 “Who Crosses the Border: A View of the SanDiego/Tijuana Metropolitan Region,” TechnicalReport, San Diego Dialogue, University of Cali-fornia, San Diego, 1993.

6 “The Economic Impacts of Mexican Visitors toArizona: 2001,” by Alberta H. Charney and VeraK. Pavlakovich-Kochi, Research Studies, Eco-nomic and Business Research Program, Univer-sity of Arizona, July 2002.

7 “The Economic Impact of Mexican Visitors tothe Lower Rio Grande Valley 2003,” by SuadGhaddar, Chad Richardson and Cynthia J.Brown, Technical Report, Center for Border Eco-nomic Studies, University of Texas–Pan Ameri-can, 2004.

8 “The Effect of the Peso Devaluation on TexasBorder Cities,” by Philip N. Diehl, Texas Busi-ness Review, vol. 57, May/June 1983, pp.120–25.

9 “The Effects of the Peso Devaluation on Cross-Border Retailing,” by J. Michael Patrick andWilliam Renforth, Journal of Borderlands Stud-ies, vol. 11, Spring 1996, pp. 25–41.

10 “The Effects of a Depreciation of the Peso onCross Border Retail Sales in San Diego andImperial Counties,” by James Gerber, workingpaper, San Diego State University, June 1999.

11 “Transportation Infrastructure and the BorderEconomy,” by Keith R. Phillips and Carlos Man-zanares, The Border Economy, Federal ReserveBank of Dallas, June 2001.

McAllen

Laredo El PasoBrownsville

’01’00’99’98’97’96’95’94’93’92’91’90’89’88’87’86’85’84’83’82’81’80’79’7850

150

250

350

450

550

650

0

.004

.008

.012

.016

.020

Real exchange rate

SOURCES: Authors’ calculations with data from Texas Comptroller of Public Accounts and Banco de México.

Real exchange rate

McAllen

Laredo

El Paso

Brownsville

’01’00’99’98’97’96’95’94’93’92’91’90’89’88’87’86’85’84’83’82’81’80’79’780

.004

.008

.012

.016

.020

-10

0

10

20

30

40

50

60

70

SOURCES: Authors’ calculations; Federal Reserve Bank of Dallas.


IIn the 1990s, the Texas economyexceeded even the remarkable perform-ance of its U.S. counterpart. State jobgrowth averaged 2.9 percent per year from1990 to 2000, well ahead of the 1.8 percentannual increases in the United States.Three engines drove the Texas economyforward in the 1990s: the oil sector, hightech (especially in Austin and Dallas) anda boom in border-city employment.Employment growth in the four largestTexas border cities topped that of thenation, and the three south Texas citiesoutperformed the state by a wide margin(Table 1).1

The accelerated job growth along theTexas–Mexico border was the result ofseveral factors: a quick Mexican recoveryafter the 1994–95 financial crisis; tightlabor markets in the United States thatattracted employers to the border insearch of the region’s surplus labor; astrong peso for much of the period, which

increased retail sales in U.S. border cities;and rapid expansion of the maquiladoraindustry.

Maquiladora expansion came on theheels of NAFTA implementation and the1994–95 peso devaluation. In recentyears, however, this part of the borderboom has turned to bust. After watching

the industry lose 290,000 jobs betweenOctober 2000 and July 2003, manyobservers are questioning the industry’sfuture. Recession, rising wages in Mexico,low-wage competition from countriessuch as China and Mexico’s inability todeal with growing problems in its compet-itive environment have all contributed to

Texas BorderEmployment and Maquiladora GrowthJesus Cañas, Roberto Coronado and Robert W. Gilmer

Table 1Percent Job Growth Along the Texas–Mexico Border

Texas El Paso Laredo Brownsville McAllen

1991 .7 2.3 4.3 2.0 1.51992 1.9 3.6 8.7 4.8 5.31993 3.3 2.2 3.6 4.4 4.51994 4.3 3.9 7.2 6.0 7.31995 2.9 0 –5.5 .4 2.71996 3.2 2.0 5.1 2.8 3.31997 4.4 2.7 7.6 3.0 3.81998 3.6 1.1 2.7 2.3 5.61999 2.2 2.0 5.3 5.8 6.52000 2.8 1.5 3.0 5.1 5.1

1990–2000 2.9 2.1 4.1 3.7 4.6

SOURCE: Federal Reserve Bank of Dallas, El Paso Branch, with data from the Texas Workforce Commission.


Cour

tesy

of M

exic

oNow

mag

azin

e


the recent downturn.This article looks at the maquiladora’s

role in today’s Texas economy, especiallyhow it affects Texas border cities. We alsoassess the industry’s future and theprospects for the maquiladora to again bea significant factor in job growth in Texasand Mexico.

Growth and DeclineThe maquiladora industry began in

1965 and experienced slow but steadygrowth under the Border IndustrializationProgram. The canceled Bracero Programhad used Mexican labor in agriculture,and the replacement maquiladora wasdesigned to relieve the resulting highunemployment rates in northern Mexico.The new program used low-wage Mexicanlabor as a lure to draw U.S. manufacturingto the region, allowing companies tomove production machinery andunassembled parts into Mexico withouttariff consequences, as long as the assem-bled product was returned to the UnitedStates for final sale.

Chart 1 shows the elevenfold increasein maquiladora employment between1980 and its peak in 2000, from 120,000workers to 1.3 million. In 1980, about 94percent of maquiladora employment wasin the border states of northern Mexico.2

Today, the share has slipped to 76 percent,but the northern states still dominate. In2004, 2,810 operating plants accountedfor about 9 percent of formal employmentin Mexico, or 3 percent of the total laborforce. The companies operating under themaquiladora program are a who’s who ofU.S. industry, including Delphi, Mattel,

Tyco, General Electric and ITT. The maquiladora industry has been

highly cyclical since its inception, fallinginto its first recession in 1974 with an 11.5percent decline in employment. Table 2shows the uneven effects of the latestmaquiladora downturn on Mexican bor-der cities. Maquiladora employment inCiudad Juárez was higher than in all theother cities combined when the recessionbegan, and it has sustained the largestpercentage losses from peak to trough(27.7 percent). Piedras Negras, NuevoLaredo and Matamoros also suffered largepercentage losses, all in excess of 24 per-cent. Ciudad Acuña and Reynosa wereexceptions to the deep recession, withCiudad Acuña declining only 10.6 percentand Reynosa continuing to grow through-out the downturn. Newer plants, a betterindustry mix and a business-friendlyenvironment account for their better per-formance.

The cyclical nature of the maquila-dora industry is not surprising, given its close ties to U.S. manufacturing (Chart

2). Throughout the latest recession andslow recovery, manufacturing was thehardest hit part of the U.S. economy, andmaquiladora output and employmentgenerally followed the lead of U.S. indus-trial production. In mid-2003, however,strong U.S. industrial growth finallyreturned, and as Table 2 shows, maqui-ladora employment has returned to re-covery as well. Job growth remainsuneven among the Mexican border cities,however, with Ciudad Acuña, Matamorosand Piedras Negras recovering moreslowly.

How Do Maquladoras Affect theTexas Border Economy?

The original vision for maquiladoraswas the “twin plant,” with capital-inten-sive operations located a few miles insidethe U.S. border and low-wage, labor-intensive operations close by on the Mex-ican side. However, the bulk of U.S. man-ufacturing was already established in theMidwest, and trucking deregulationwould make transportation links betweenthe border and the Midwest both easierand cheaper in the 1970s and ’80s. Thetwin-plant vision was never realized alongthe border. Instead, the maquiladora sup-ply chain remained concentrated in statessuch as Illinois, Michigan and Ohio.

What economic impact would a newmaquiladora in Mexico have on a neigh-boring U.S. city? The list might run as fol-lows. To select and develop a site, U.S.legal, engineering and financial assistancewould be used. Once established, the newplant would rely on U.S.-based businessesfor customs, brokerage, warehousing andtransportation services. The plant wouldalso purchase a variety of office, packag-ing and industrial supplies. Corporatemanagement, engineers and quality spe-

Chart 1Maquiladora Employment Growth

Thousands, seasonally adjusted data

0100200300400500600700800900

1,0001,1001,2001,3001,400

’04’02’00’98’96’94’92’90’88’86’84’82’80 ’05’03’01’99’97’95’93’91’89’87’85’83’81

SOURCES: Federal Reserve Bank of Dallas, El Paso Branch, with data from INEGI.

Table 2Texas–Mexico Maquiladora Border Employment

Peak TroughJobs Date Jobs Date April 2005

Ciudad Juárez 262,550 October 2000 189,930 June 2003 213,389Ciudad Acuña 37,512 November 2002 33,541 February 2005 33,674Piedras Negras 15,222 February 2000 10,939 December 2004 11,187Nuevo Laredo 22,915 February 2000 17,171 April 2003 22,233Reynosa 86,925 April 2005 N/A N/A N/AMatamoros 68,413 October 2000 51,900 August 2003 53,002

NOTES: Seasonally adjusted data; border twin cities are as follows: Ciudad Juárez–El Paso, Ciudad Acuña–Del Rio, Piedras Negras–EaglePass, Nuevo Laredo–Laredo, Reynosa–McAllen and Matamoros–Brownsville.

SOURCE: Federal Reserve Bank of Dallas, El Paso Branch, with data from INEGI.


research suggests that U.S. borderstates—with the exception of Arizona,where job losses ranged from negligible tosmall—gained jobs as a result of growth inthe maquiladora industry.8

A more recent development has beenthe arrival of component parts and mate-rial suppliers in U.S. border cities. Specificexamples can be found in El Paso, neigh-bor to Ciudad Juárez, which is home to thelargest number of maquiladora employ-ees along the U.S. border. Over the pastdecade, an increasing number of rubberand plastics, electronics and electricalequipment, and metal fabricating plantshave begun to operate in El Paso to serveas suppliers to the maquiladora industry(Chart 3).

Components supplied include com-puter housings, electrical wiring har-nesses, special dies and tools, and electri-cal switches. About 26 plastic-injectionmolding plants can be identified, 31 metalstamping companies, and 12 electric- andelectronic-related companies. Together,these companies employed 4,000 workersin 2004. The manufacturing sectors thatsupply the maquiladoras paid about 40percent more in hourly wages than thelow-wage apparel, textile and leather in-dustries that traditionally operated in El Paso.

Maquila manufacturing in Mexicoalso positively influences El Paso’semployment in transportation, realestate, and legal and accounting services(Chart 4). Given the rapid increase intrade flows after 1993, transportation and

warehousing employment acceleratedquickly. Business service employment,especially personnel supply services,computer programming and data pro-cessing, grew 45 percent from 1990 to2004. El Paso’s maquiladora-related busi-nesses rely heavily on temporary staffingagencies to hire additional personnel tomeet rising demand. Computer program-ming and data service workers help mini-mize the burden of paperwork required bycustoms agencies to export or importcomponents. Legal employment grew 20percent over the same period. Similarresults can be found up and down the U.S.border.

The definitive study on the linkagesbetween maquiladoras and the bordereconomy, by Gordon Hanson, takes allthese factors into account.9 Hanson esti-mates that a 10 percent increase inmaquiladora output in a Mexican bordercity will increase employment in its U.S.city pair by 1.1 to 2 percent. He providesmore specifics by estimating that thissame 10 percent increase in output wouldincrease wholesale trade employment inthe U.S. city by 2.1 to 2.7 percent, trans-portation services by 1.7 to 2.7 percent,manufacturing by 1.2 to 2.1 percent andretail trade by 1 to 1.8 percent.

The Role of RecessionThe recent recession has played an

important role in the latest downturn ofthe highly cyclical maquiladora industry.At the same time, maquiladoras have longserved as a low-wage platform for U.S.

cialists would be drawn to the border tovisit this plant, and they would spendmoney on food and lodging.3 Maquiladoraemployees draw their salary in Mexico butdo a significant share of their shopping inthe United States, stimulating employ-ment in local retail and service sectors.

These impacts on the U.S. borderhave been recognized for some time, anda number of studies were conducted inthe 1970s and ’80s to quantify them. Forinstance, in 1972, Ladman and Poulsenfound that Agua Prieta, Sonora, maquila-dora workers spent 40 percent of theirwages in Arizona.4 Ayer and Layton esti-mated the maquiladoras’ impact on valueadded and population, using an input–output model for the Arizona–Mexicoborder economy. They concluded thatMexicans’ expenditures due to the growingpresence of twin plants increased valueadded by 14 percent and population by 11percent on the U.S. side of the border.5

In a 1984 study of the Texas border,Holden estimated that maquiladoraemployment had a large impact onemployment in the border communitiesof El Paso, Laredo, McAllen and Browns-ville. For instance, a 10 percent increase inmaquiladora payroll results in a 2 to 3 per-cent increase in employment in El Pasoand McAllen as well as a 3 to 4 percentincrease in Laredo and Brownsville.6

In another study, Sprinkle found thatduring the early 1980s Ciudad Juárezmaquiladoras accounted for one of fivejobs created in El Paso, and these new jobswere concentrated in the service sector.7

Silvers and Pavlakovich assessed the rela-tive magnitude of employment gains andlosses across U.S. border regions due tomaquiladora industry activity. Their

Chart 2Maquiladora Ties to U.S. Industrial Sector

Index, January 2000 = 100, seasonally adjusted

SOURCES: Federal Reserve Bank of Dallas, El Paso Branch, with data

from INEGI; Federal Reserve Board.

Maquiladoraemployment

U.S. industrialproduction

80

85

90

95

100

105

110

20042003200220012000 2005

Chart 3Maquiladora Suppliers in El Paso

Index, 1990 = 100

Electrical equipment

Metalworking

Plastics

0

50

100

150

200

250

300

350

2002200120001999199819971996199519941993199219911990

Apparel

20042003

NOTE: North American Industry Classification System (NAICS) codes used in the chart were 3327, 3329, 3328, 3332, 3335, 3159, 3261 and 3344.

SOURCE: Federal Reserve Bank of Dallas, El Paso Branch, with data from Bureau of Labor Statistics.


manufacturing, and the rise of new low-wage alternatives such as China, Indiaand Vietnam has broadened U.S. optionsfor manufacturing. These very low-wagecompetitors, plus rising real Mexicanwages, have become a factor in pushingsome maquiladora activity abroad. Mex-ico generally has looked at the loss of thelowest wage jobs as an inevitable price ofprogress, because increasing domesticwage levels must be seen as a positiveaspect of economic development. Thegovernment has expressed reluctance toenter into subsidy programs to retain orattract these industries, considering suchaction as poor fiscal policy and a violationof Organization for Economic Coopera-

tion and Development and World TradeOrganization rules.

To focus on the question of howmaquiladoras will respond to economicrecovery and which sectors would benefit, we developed some econometricestimates. As in other models, ourmethodology confirmed that past maqui-ladora employment has primarily beendriven by the business cycle and relativereal wages.10 Trends and dummy shiftvariables were included to account forstructural change, particularly testing forbreaks with the 1994 implementation ofNAFTA and the 1994–95 financial crisis inMexico. The general methodology followsseveral papers by Branson and Love, and

detailed results are reported elsewhere.11

To examine the future of the industryunder various assumptions, we simulatedmaquiladora employment following itssecond quarter 2000 peak. The base casewas the actual outcome through 2002, adecline of 14.5 percent for the industry asa whole. Scenario 1 (S1) assumed norecession and that the U.S. unemploy-ment rate held firm at a historically low 4percent through the end of the period.Real relative wages rose in this scenario,just as in the base case. Scenario 2 (S2)assumed the recession occurred but thatreal relative maquiladora wages fell 6.1percent after second quarter 2000 insteadof rising 16.8 percent. And scenario 3 (S3)assumed the best of both worlds formaquiladora managers, falling real rela-tive wages and no recession.

Chart 5 shows the results for allmaquiladoras combined. This can becomputed two ways: as the result of a sin-gle estimate based on the sum of allmaquiladora employment or as the sumof the simulation results for 10 maqui-ladora sectors. Fortunately, they agreequite closely. Eliminating the U.S. reces-sion in S1 would provide an increase ofapproximately 20 percent in employmentin the simulation period, replacing adecline of about 14.5 percent in the basecase. The percentage turnaround for S2 issimilar, and the combined effect in S3 is a31 percent increase.

Four individual sectors do not

Chart 5Simulation Results

NOTE: The base cases in the two calculations are slightly different because the chemicals sector was excluded from the sum of regression results. There was a break in the data for this sector.


-14.6

20.1

17.7

31

–20 –16 –12 –8 –4 0 4 8 12 16 20 24 28 32 36

Base

Scenario 1

Scenario 2

Scenario 3

–20 –16 –12 –8 –4 0 4 8 12 16 20 24 28 32 36

–14.5

21.3

31.3

Base

Scenario 1

Scenario 2

Scenario 3

21.3

Total Maquiladoras Sum of Maquiladora Sectors

Percent change in jobs Percent change in jobs

Chart 4Service Employment in El Paso

Index, 1990 = 100

Legal and accounting services

Real estate

2002200120001999199819971996199519941993199219911990

Transportation andtransportation services

200420030

50

100

150

200

250

300

350

NOTE: North American Industry Classification System (NAICS) codes used in the chart were 4889, 5311, 5312, 5313, 5411 and 5412.

SOURCE: Federal Reserve Bank of Dallas, El Paso Branch, with data from Bureau of Labor Statistics.


respond to an upturn in the U.S. econ-omy: leather, toys, furniture and a groupof other, unclassified maquiladoras. Theirsimulation results are summarized inTable 3. Combined, these maquiladorasaccounted for 226,782 jobs at the secondquarter 2000 peak, or 18.1 percent of thetotal. These sectors are unlikely to returnto growth with U.S. economic recovery.

The three largest maquiladora sec-tors, together accounting for 76.1 percentof the peak employment, all respond pos-itively to economic recovery in the simu-lations. In S1, electrical machineryrecords an 18.2 percent increase, in placeof a 26.1 percent decline. Textiles turnaround to record a 63.2 percent gain in S1,and transportation equipment (which didnot decline after second quarter 2000)grows by another 4.5 percent in this sce-nario.

In conclusion, less than 20 percent ofmaquiladora employment is in sectorsthat are unresponsive to economic recov-ery in the United States, and overallgrowth seems likely to continue. However,even those sectors that continue to growin simulations are going to be influencedby foreign competition. The effect of for-eign competition is often couched interms of a product cycle, in which productdevelopment and testing occur in theUnited States, initial long production runstake place in Mexico and ultimately prod-uct commoditization happens in China oranother low-wage competitor. The morequickly and easily a product is commodi-tized, the quicker it will move to China.Leather, toy and furniture sectors areoften cited as no longer competitive inMexico. But even within the mostadvanced sectors, we may find individualproducts susceptible to being lost to lowerwage countries in exactly the same way—computers, cell phones, modems, print-ers and disk drives, for example. Hence,

the rise of foreign competition meanseven sectors returning to positive growthwith economic recovery may experienceslower job growth than in the recent past,as some products within the sector arecommoditized.

In assessing Mexico’s competitiveprospects, the nation retains crucialadvantages over the rest of the world,even as domestic wages rise. The mostimportant factor is proximity to the U.S.market. For example, bulky items thathave a high ratio of weight to value, suchas large-screen televisions or major appli-ances, will remain competitive. Proximityalso matters if the inventory cycle is short,if there are constant design changes or ifthere must be frequent retooling. Mexicowill also be competitive when quality ismore important than price, such as withmedical equipment or when intellectualproperty rights are critical.12

Texas-Based SuppliersThe maquiladoras’ contribution to

U.S. border city growth in the 1990sstemmed from (1) the spillovers fromrapid maquiladora expansion in neigh-boring Mexican cities and (2) the shift ofmany maquiladora suppliers to bordercities from their base in the Midwest. Wehave already shown how foreign competi-tion and rising real wages in Mexico havereduced the prospects for maquiladoragrowth, but foreign competition is alsomaking significant inroads into themaquiladora supply chain. This raises thepossibility of slowing, or even reversing,the increase of U.S. border-city suppliersto the maquiladora industry.

Throughout the 1990s, the UnitedStates supplied the vast majority ofmaquiladora industry inputs. In 2000, 90percent of maquiladora inputs were fromthe United States and 9 percent were fromAsia, with China contributing only 1 per-cent (Chart 6). By 2004, 59 percent camefrom the United States and 35.7 percentfrom Asia, including 11.1 percent fromChina. The United States remains themajority supplier, but this rapidly movingtrend continued to run in favor of Asiainto 2005.

The vehicle for entry of foreign inputsto Mexico is 20 sectoral promotion pro-grams, or PROSECs, created by the Mexi-can government in December 2000. Theywere created in response to implementa-tion of NAFTA Article 303, which in Janu-ary 2001 eliminated duty-free imports ofmaquiladora inputs from non-NAFTAcountries. The PROSECs protect the entryto Mexico of non-NAFTA componentsthat are not readily available in thedomestic market, allowing them to enterunder reduced tariffs of zero to 5 percent.Despite the paperwork and the need totrack the origin of thousands of parts tocomply with PROSECs, maquiladorashave apparently fully embraced the pro-grams.

Data are not available on exactlywhich inputs are being displaced, makingit difficult to assess the impact on Texasborder communities. For example, if the1990s shift of suppliers to the border fromthe Midwest was based on just-in-timeinventory needs, it may be difficult forAsian suppliers to take their place. How-ever, given the extent and pace at which

Table 3Maquiladora Sectors That Are Unresponsiveto a U.S. Economic Rebound(Percent change in jobs)

NotLeather Toys Furniture classified

Base –25.6 –31.5 –8.7 –4.9S1 –7.8 –21.0 –8.9 –2.1S2 –6.8 –34.7 17.3 –22.1S3 –7.4 –5.1 4.4 5.8


Chart 6Mexican Maquiladora Imports by Country of Origin

Percent

0

10

20

30

40

50

60

70

80

90

100

OtherChinaAsia without ChinaUnited States

20022003

20002001

20042005:Q1

SOURCE: Federal Reserve Bank of Dallas, El Paso Branch, with data from Banco de México.


Asian suppliers have taken market share,it would be hard to argue that themaquiladora market share of Texas-basedsuppliers has not been reduced. Futureexpansion of Texas-based suppliers islikely to slow as well.

ConclusionMexico’s maquiladora jobs are grow-

ing once more, beginning with theresumption of U.S. industrial expansionin mid-2003. Mexico retains importantcompetitive advantages over many of itslow-wage rivals, based on proximity to theUnited States, political and financial sta-bility, and the rule of law. The maqui-ladora industry is stable, competitive andgrowing again.

It is unlikely, however, to repeat thebanner performance of the 1990s, at leastnot in the near future. There were ele-ments of unique, one-time stimulus in the1990s, with the collapse of the peso in1994–95 and the implementation ofNAFTA in 1994. Further, foreign competi-tion appears to have taken away thepotential for any growth in several low-wage sectors and probably has reducedthe growth potential of a number of othersectors as well.

Rising real wages in Mexico haveaccelerated the transfer of low-wage jobsto other countries, and the Mexican gov-

ernment has argued that this must beseen as a highly desirable result of suc-cessful economic development and Mex-ico’s move up the product cycle. The nextgeneration of maquiladoras should not bejudged by the ability to generate low-wagejobs, but by productivity, value added orrising wages. Critics, at the same time,claim Mexico simply has not done an ade-quate job of preparing the way for moresophisticated manufacturing. To illustratethis point, many observers cite the failure(so far) of proposed reforms in energy,labor law, taxes and telecommunications.These and other reforms are badly neededto prepare Mexico for a fine market econ-omy.

Finally, it is not just the maquiladoraindustry that is affected by foreign com-petition, but the U.S.-based supply chainas well. In 2000, 90 percent of inputs to themaquiladoras came from the UnitedStates, and four years later that numberwas only 59 percent. Texas border cities inthe 1990s developed rapidly as a critical,new part of this supply chain, with suppli-ers shifting from the Midwest to theU.S.–Mexico border. We lack industrydetail to know exactly how the recent suc-cess of foreign suppliers is affecting Texasborder cities, but again, declining eco-nomic stimulus from maquiladora expan-sion would seem to be the rule.

Cañas and Coronado are assistanteconomists at the El Paso Branch of theFederal Reserve Bank of Dallas. Gilmer is avice president at the Federal Reserve Bankof Dallas.

Notes1 The four border cities contributed 6.8 percent of

the 2.3 million jobs generated in Texas from1990 to 2000, while making up 6.1 percent ofTexas employment in 1990.

2 Mexican border states include Baja California,Sonora, Chihuahua, Coahuila and Tamaulipas,excluding Nuevo León.

3 “The Employment Impact of MaquiladorasAlong the U.S. Border,” by J. Michael Patrick, inThe Maquiladora Industry: Economic Solution orProblem?, ed. Khosrow Fatemi, New York:Praeger Publishers, 1990, pp. 31–35.

4 “Economic Impact of the Mexican Border Indus-trialization Program: Agua Prieta, Sonora,” byJerry R. Ladman and Mark O. Poulsen, ArizonaState University, Center for Latin American Stud-ies, May 1972.

5 “The Border Industry Program and the Impact ofExpenditures on a U.S. Border Community,” byHarry Ayer and Ross Layton, Annals of RegionalScience, vol. 8, 1974, pp. 105–17.

6 “Maquiladoras Along the Texas–Mexico Border:An Econometric Evaluation of Employment andRetail Sales Effect on Four Texas BorderSMSAs,” by Richard J. Holden, Texas Depart-ment of Community Affairs, Regional EconomicDevelopment Division, February 1984.

7 “Project Link: An Investigation of EmploymentLinkages Between Cd. Juárez and El Paso,” byRichard Sprinkle, University of Texas at El Paso,December 1986.

8 “Maquila Industry Impacts on the Spatial Redis-tribution of Employment,” by Arthur L. Silversand Vera K. Pavlakovich, Journal of BorderlandsStudies, vol. 9, December 1994, pp. 47–64.

9 “U.S.–Mexico Integration and RegionalEconomies: Evidence from Border-City Pairs,”by Gordon Hanson, Journal of Urban Econom-ics, vol. 50, September 2001, pp. 259–87.

10 For more information, see the papers from theDallas Fed conference “Maquiladora Downturn:Structural Change or Cyclical Factors?” available atwww.dal lasfed.org/news/research/2003/03maquiladora.html. In particular, see the pre-sentations by Everardo Elizondo Almaguer,Banco de México; William C. Gruben, FederalReserve Bank of Dallas; James Gerber, SanDiego State University; and Ernesto AcevedoFernández, Ministry of Finance and PublicCredit.

11 “Maquiladora Downturn: Structural Change orCyclical Factors?” by Jesus Cañas, RobertoCoronado and Robert W. Gilmer, InternationalBusiness and Economics Research Journal, vol.3, August 2004; “Dollar Appreciation and Manu-facturing Employment and Output,” by WilliamH. Branson and James P. Love, National Bureauof Economic Research, Working Paper No.1972, July 1986; “The Real Exchange Rate andEmployment in U.S. Manufacturing: State andRegional Results,” by William H. Branson andJames P. Love, National Bureau of EconomicResearch, Working Paper No. 2435, November1987; “The Real Exchange Rate, Employmentand Output in Manufacturing in the U.S. andJapan,” by William H. Branson and James P.Love, National Bureau of Economic Research,Working Paper No. 2491, February 1988.

12 See “Maquiladora Downturn: Structural Changeor Cyclical Factors?” by Jesus Cañas, RobertoCoronado and Robert W. Gilmer, FederalReserve Bank of Dallas Business Frontier, Issue2, 2004.

Mexico retains

important

competitive

advantages over

many of its low-wage

rivals, based on

proximity to the

United States,

political and

financial stability,

and the rule of law.

T


Texas and oil. These two words havegone hand in hand since 1889, when thestate started producing oil. Since then, theTexas economy has often been driven byvolatile energy prices—suffering with lowoil prices and benefiting with high oilprices.

The effects of energy prices on theTexas economy were particularly evidentduring the 1970s and 1980s (Chart 1). Asenergy prices rose, the Texas economyexpanded at a rapid pace, with strongemployment and income growth.Although the Texas economy continued toexpand until 1986, the oil and gas sectorbegan to slip as energy prices slid fromtheir 1981 heights. The oil price collapsein July 1986 touched off a statewide reces-sion and significant job losses.

Since the early 1980s, however, theTexas energy industry has shrunk andother sectors of the Texas economy havegrown. Despite these changes, Texasremains the top oil and natural gas pro-ducer in the United States and exportsmost of its production of these two com-modities to other states. Consequently,the energy industry remains an importantdriver of the state economy.

The diversification of the Texas econ-omy away from energy and this sector’scontinuing importance to the stateprompt us to consider: How much doswings in energy prices affect the Texaseconomy today? How much has that rela-tionship changed since the energy boomyears of the 1970s and 1980s?

Oil Production in Texas: A Brief History1

The first economically significant oilin Texas was discovered in Corsicana in1894. Discoveries in Navarro County fol-lowed. By 1901 the Spindletop oil field wasproducing 75,000 barrels per day and hadcontributed to the first Texas oil boom.

In the early 1900s, Texas produced rel-atively little oil and gas—crude oil pro-

Do Higher Oil Prices Still Benefit Texas?Stephen P. A. Brown and Mine Yücel


duction was only about 1.3 percent oftotal U.S. production, and natural gas was0.1 percent of U.S. production. By 1952,Texas’ shares of total U.S. crude oil andnatural gas production peaked at 45 and52.2 percent, respectively. Crude oil andnatural gas production continued toincrease in the state, with the peak forboth coming in 1972.

As oil and gas production increased inTexas, so did their importance to the stateeconomy. The creation of OPEC in 1960and subsequent oil price increases in the1970s and early 1980s gave rise to a boomin the Texas economy. Oil and gas outputbecame an increasing share of Texas out-put (Chart 2). In 1981, at the height ofworld oil prices, oil and gas extraction wasabout 20 percent of total Texas gross stateproduct.

After reaching $38 per barrel in 1981,oil prices began softening. Gradually slid-ing during the next few years, pricesfinally collapsed to $11.82 per barrel inJuly 1986. This led to a recession in Texasthat lasted 17 months and had a devastat-ing effect on state employment.

The number employed in the Texasmining industry (which is mostly oil andgas extraction) rose from about 7,000 in1900—0.7 percent of total state employ-ment—to 90,000 by 1950—a 3.1 percentshare. At the oil and gas industry’s peak in1981, Texas employment in oil and gasextraction and oilfield machinery reached366,200 — 6 percent of total nonfarmemployment in the state (Chart 3). By thetime the oil industry bottomed out in1987, 175,000 jobs had been lost in the oiland gas extraction and oilfield machinerysectors.

Refining and PetrochemicalsAfter the first Texas refinery opened in

the Corsicana oil field in 1898, the petro-leum refining and petrochemical indus-tries flourished in the state. In 1939 (theearliest data available from the U.S. Cen-sus of Manufacturers), the chemicalindustry employed about 6,800 produc-tion workers, and the petroleum refiningindustry employed 19,000 (accounting for5.5 and 15 percent of total manufacturingemployment, respectively). Refining’sshare of state output was highest in 1939at 28 percent of total manufacturedgoods. By 1958, the Texas petroleum refin-ing industry reached its zenith with 43,000

Chart 1Texas Employment Tracks Oil Prices in 1970s and 1980s

Employment (in thousands) 2004 dollars per barrel

Refiners' acquisition cost

Detrended Texasemployment

–600

–400

–200

0

200

400

600

800

’01’98’95’92’89’86’83’80’77’740

10

20

30

40

50

60

70

80

90

SOURCES: Bureau of Labor Statistics; Federal Reserve Bank of Dallas; Energy Information Administration; authors’ estimates.

Chart 2Oil and Gas Extraction’s Share of Texas Output Peaks in 1981

Percent of Texas GSP

0

5

10

15

20

’01 ’99 ’97 ’95 ’93 ’91 ’89 ’87 ’85 ’83 ’81 ’79 ’77 ’75 ’73 ’71 ’69 ’67 ’65 ’63

SOURCE: Federal Reserve Bank of Dallas.

Chart 3Energy Sector Employment Declines After Early 1980s

Share of Texas nonfarm employment (percent)

0

.5

1

1.5

2

2.5

3

3.5

4

4.5

5

’01’99’97’95’93’91’89’87’85’83’81’79’77’75’73’71’69’67

Chemicals and allied products

Petroleum and coal productsOilfield machinery

Oil and gas extraction



employees.Today, the refining industry con-

tributes about 11 percent of Texas manu-facturing output and 1.5 percent of totalTexas output. Employment has alsosteadily declined to less than 0.3 percentof total Texas employment (Chart 3). Thepetrochemical industry provides about 12percent of Texas manufacturing output,1.6 percent of total Texas output and lessthan 0.9 percent of total Texas employ-ment.

The refining and petrochemicalindustries provide some counterbalanceto the effects of changing energy prices onthe Texas economy. These two industriesgenerally are hurt by rising oil and naturalgas prices.

Diversification of the Texas Economy

As output in the Texas mining indus-try shrank, output in other Texas indus-tries continued to grow after the mid-1980s. Texas saw output gains in manu-facturing, construction, agriculture andthe service-producing sectors — whole-sale and retail trade; transportation, com-munications and public utilities (TCPU);services; finance, insurance and realestate (FIRE); and government (Chart 4).Growing at a faster rate than total Texasgross state product, manufacturing, trade,TCPU, services and FIRE accounted forincreasing shares of Texas output. In con-trast, agriculture, construction and gov-ernment posted decreasing shares.

A similar picture emerges for Texasemployment since the mid-1980s. Ser-vices, construction and trade grew fasterthan total employment and accounted forincreasing shares of Texas nonfarmemployment (Chart 5 ). Employmentshares for TCPU and FIRE remained rela-tively constant, while those for manufac-turing and government decreased alongwith mining.

Oil and the Texas EconomyEven without a rigorous analysis, it’s

evident the relationship between energyprices and the Texas economy haschanged since the 1980s. Oil and gas pro-duction accounted for 19.4 percent ofTexas output in 1981 and only 6 percent in2002. Similarly, output and employmentin energy-related industries, such as oiland gas field machinery, claim a smaller

share of the Texas economy today than inthe early 1980s.

To examine in more detail how theTexas economy’s diversification awayfrom energy-producing industries hasaffected its response to volatile energyprices, we developed an econometricmodel that captures the effects of oil priceshocks on the Texas economy for theperiod 1970–2002.2 We find that the rela-tionship between oil prices and the Texaseconomy is considerably different todaythan it was during the oil boom and bustyears of the 1970s and 1980s.

Our analysis reveals that the relation-ship between oil prices and the Texaseconomy breaks between 1987 and 1988,which indicates that the effects of chang-ing oil prices on the economy were differ-ent in 1970–87 than in 1988–2002. To

determine just how this relationship dif-fered across the two periods, we analyzethe data in two different ways. We exam-ine how much of the actual fluctuation inTexas output and employment arose fromoil price shocks and other causes in eachof the two periods. We also estimate andcompare by how much Texas output andemployment would have responded to a10 percent oil price shock in each of thetwo periods.

We find changes in oil pricesaccounted for a much higher percentageof fluctuations in the Texas economy in1970–87 than in 1988–2002. In the earlierperiod, nearly half the fluctuation in Texasoutput (46 percent) arose from changingoil prices. In the latter period, however,less than 10 percent of Texas output fluc-tuations arose from oil price shocks. In

Chart 4Texas Economy Diversifies Away from Mining After Mid-1980s

GSP index, 1971 = 100

0

100

200

300

400

500

600

’01’99’97’95’93’91’89’87’85’83’81’79’77’75’73’71

TradeTCPUServicesManufacturingFIRE

TotalAgricultureConstructionGovernmentMining

NOTE: TCPU is transportation, communications and public utilities; FIRE is finance, insurance and real estate.


Chart 5Texas Employment Shifts Away from Mining After Early 1980s

Texas Employment Index, 1970 = 100

75

125

175

225

275

325

375

425

475

’02’00’98’96’94’92’90’88’86’84’82’80’78’76’74’72’70

ServicesTrade TCPU

ManufacturingFIRETotal

Construction

Government

Mining

NOTE: TCPU is transportation, communications and public utilities; FIRE is finance, insurance and real estate.



contrast, the fluctuations in U.S. GDPaccounted for about 40 percent of thefluctuations in Texas output in the latterperiod.

The Response to Oil Price Shocks

The Texas economy’s response to anoil price shock is significantly different inthe two periods (Table 1). For 1970–87, weestimate that an oil price increase wouldhave led to sustained gains in both outputand employment. In particular, a 10 per-cent increase in oil prices would have ledto a 2.6 percent increase in Texas grossstate product and about a 1 percentincrease in employment.3 An oil priceincrease of 10 percent also would havetemporarily boosted the growth rate ofthe Texas economy, with output growing 1percent faster during the next few quar-ters and employment growing 0.1 percentfaster over the next three to four months,then a little slower thereafter.

The economy was much less respon-sive to oil prices in the period 1988–2002,and the nature of the response was differ-ent. In the second period, a 10 percentincrease in oil prices would have led toonly about a 0.4 percent gain in gross stateproduct. The net response of employmentto a rise in oil prices is basically nil. Thenegligible result in employment may arisefrom the energy sector’s greatly mutedresponse to oil price fluctuations in thelatter period and the inability or reluc-tance of oil companies to hire newemployees as energy prices rose.

To further examine the channelsthrough which oil price shocks affect theTexas economy, we examined the effectsof oil price shocks on the rig count and oiland gas employment in both periods. Wefound that the rig count responded muchmore strongly to oil price increases in thefirst period than in the second. For1970–87, we estimate that a 10 percentincrease in oil prices would have boostedthe rig count by 20 percent. In contrast,the same percentage increase in oil prices

in 1988–2002 would have yielded only a6.6 percent increase in the rig count.

Similarly, oil and gas employmentshowed a much smaller response in thesecond period. We estimate that a 10 per-cent increase in oil prices would have gen-erated a 9.5 percent increase in Texas oiland gas employment for 1970–87 but onlya 1.1 percent employment increase in1988–2002.

One reason for the weaker responsein the rig count and employment may bechanges in technology. After the 1986crash in oil prices, companies improvedoilfield technology and produced more oilwith fewer rigs. Therefore, the same rise inoil prices brings forth fewer rigs and oil-field workers in the latter period. In addi-tion, contacts in the industry say there arefewer prospects for new drilling in Texas,and companies are increasingly shiftingtheir drilling overseas.4

Oil Price Effects on the Texas Economy

Over the past 20 years, the Texasenergy industry has shrunk while othersectors of the Texas economy have grown.Nonetheless, Texas produces more oil andgas than any other state in the nation.Texas accounts for 20 percent of crude oiland 26 percent of natural gas productionin the United States (excluding federal off-shore). Texas also exports oil and naturalgas to the rest of the nation. Conse-quently, higher energy prices still benefitthe state—even if it is by less than in theboom years of the 1970s and early 1980s.

Our estimates confirm the Texaseconomy has become less sensitive to oilprice fluctuations, but it still respondsfavorably to higher energy prices. Duringthe 1970–87 period, a 10 percent increasein oil prices would have boosted Texas

gross state product by 2.6 percent andemployment by 1 percent. During the1988–2002 period, a 10 percent increasein oil prices would have raised Texas grossstate product by 0.4 percent with no sig-nificant net effect on employment.

We find evidence for two ways inwhich the Texas economy has becomeless sensitive to fluctuations in oil pricesthan it was in the 1970s and 1980s. Thefirst is that oilfield activity has becomeless sensitive to fluctuations in energyprices. The second is that the energyindustry makes up a smaller share of theTexas economy than it used to. Togetherthese factors have meant that Texas out-put is about 15 percent as sensitive to oilprice fluctuations as it was from 1970 to1987. Texas nonfarm employment nolonger seems to be affected by oil pricefluctuations.

Brown is a senior economist and assistantvice president and Yücel is a senioreconomist and vice president in theResearch Department of the FederalReserve Bank of Dallas.

NotesThis article was previously published under thetitle “The Effect of High Oil Prices on Today’sTexas Economy,” in Federal Reserve Bank ofDallas Southwest Economy, September/October2004.

1 See “Oil and Gas Industry,” The Handbook ofTexas Online, www.tsha.utexas.edu/handbook/online.

2 We use a vector-autoregressive model with oilprices, U.S. GDP, Texas gross state product,Texas nonfarm employment, Texas employmentin oil and gas extraction, and the Texas rig countas variables.

3 These results are similar to those found in“Energy Prices and State Economic Perfor-mance,” by Stephen P. A. Brown and Mine K.Yücel, Federal Reserve Bank of Dallas EconomicReview, Second Quarter 1995. Using input–out-put analysis, Brown and Yücel estimate that a 10percent increase in oil prices would haveboosted Texas employment by 1.37 percent in1982 and by 0.3 percent in 2000.

4 Drilling has shifted toward natural gas in theUnited States and Texas, but because naturalgas prices generally moved with oil prices dur-ing the estimation periods, the shift may notalter the rig count’s weakening response to oilprices.

Table 1Effect of a 10 Percent Increase in Oil Prices on Texas Economy

Texas OilTexas nonfarm Rig and gasGSP employment count employment

1970–1987 +2.6% +1.0% +20% +9.5%1988–2002 +0.4% 0 +6.6% +1.1%

Our estimates confirm

the Texas economy has

become less sensitive

to oil price fluctuations,

but it still responds

favorably to higher

energy prices.

R


Rich natural resources, abundantland, a central location within the UnitedStates and a business-friendly environmenthave long attracted both immigrants andU.S. natives to Texas. As a result, the state’spopulation is faster growing, younger andmore diverse than the nation’s.

These rapid demographic changespresent challenges for the future. As thestate’s baby boomer population ages,more demands will be placed on housing,health care and social services. Hispanics,already a dominant force in Texas, areexpected to become the majority popula-tion group by 2020. The significantincrease in this population (both immi-grant and native) has far-reaching impli-cations for education, housing and thelabor force. The key issue facing Texas willbe to reduce the economic and educa-tional disparities prevalent among thestate’s ethnic groups as the populationcontinues to grow and evolve.

This article looks at populationgrowth and demographic changes ofrecent decades. Then, with projectionsfrom the Texas State Data Center, weexamine some sectors of the economythat will be challenged by these demo-graphic forces in the coming decades.

Texas: Big and Getting BiggerSince the early 1900s, Texas has grown

faster than the nation. However, duringthe Texas oil boom, the state’s populationgrowth accelerated. From 1970 to 1980, asoil prices spiraled upward and peopleflocked to Texas, its population grew by2.71 percent per year, while the nation’sincreased at a 1.14 percent pace (Chart 1).Even during the 1980s, which witnessedan oil and real estate bust, Texas almostdoubled the nation’s population growth.

During the 1990s, Texas againexceeded expectations and grew by itslargest amount yet, adding almost 3.9 mil-lion residents and surpassing New York asthe second most populous state. Manyimmigrants and residents from other

The key issue facing Texas will be to

reduce the economic and educational

disparities prevalent among the state’s

ethnic groups as the population

continues to grow and evolve.

The Changing Face of TexasPopulation Projections and ImplicationsD’Ann Petersen and Laila Assanie


Table 2Total Population and Components of Population Change in United States and Texas

NaturalNatural increase Net migration increase Net migration

Totalpopulation Net Net

change Net Net Total international internalGeographic (April 2000– international internal Total change migration migration

area July 2003) Births Deaths Total migration migration Total (percent) (percent) (percent) (percent)

United States 9,364,374 13,098,788 7,843,040 5,255,748 4,108,626 0 4,108,626 56.12 43.88 43.88 0Texas 1,259,945 1,189,400 489,715 699,685 430,048 130,212 560,260 55.53 44.47 34.13 10.33

SOURCE:Census Bureau.

states were drawn to Texas’ strong econ-omy and rapidly expanding high-techcenters, such as Austin and Dallas’ tele-com corridor.

Even with the drastic economicdownturn of 2001, which hit Texas muchharder than most other areas of thenation, the state gained an additional 1.26million residents from 2000 through 2003,for a total of 22 million, again growingtwice as fast as the nation. Althoughdomestic in-migration—people movingto Texas from other states within theUnited States—slowed during Texas’ hardeconomic times, the state’s high birthrateand a strong pace of immigration keptpopulation growing at a healthy speed.The combination of these factors —higher international immigration, a highHispanic birthrate and less domesticmigration—resulted in Texas’ Anglo pop-ulation dipping below the majority levelof 50 percent in 2003 for the first timesince the 1800s.

Why the Rapid Growth? Two major factors are spurring Texas’

rapid population growth. One is the state’s

higher-than-average birthrate. This ispartly a result of the state’s Hispanic her-itage and its ties to Mexico, where totalfertility rates were 2.5 percent in 2004,quite a bit higher than the United States’2.1 percent.1 In 2000, Texas was second inthe country (behind Utah) in state rank-ings for birth/fertility rates. Because birth-rates change slowly over time, Texas willprobably continue to see large naturalincreases in its population despitechanges in economic conditions or immi-gration policies.

Perhaps the most important factorbehind Texas’ more recent populationgrowth is the strong pace of net migration.Historically, people have been drawn toTexas because of its abundant land andnatural resources. In more recent years,people and businesses were drawn byTexas’ robust economy and favorablebusiness climate. Net migration, whichincludes both domestic in-migration andinternational immigration, was highestduring periods of greatest economicexpansion—the 1970s oil boom (58.4 per-cent) and the 1990s high-tech/telecomboom (50.4 percent)—and accounted for

a larger share of the state’s populationgrowth than natural increase (Table 1).Interestingly, even with the state’s reces-sion in 2001–03, net migration remainedrelatively high, thanks to strong interna-tional immigration, accounting for 44.5percent of Texas’ population increase.

How Has Immigration Changedthe Face of Texas?

The healthy pace of Texas’ populationgrowth that began in the 1990s is due inlarge part to strong international immi-gration, which surpassed domestic in-migration as a contributor to populationgrowth in six of the nine years during the1990s.2 Immigration reached historic pro-portions as the number of foreign-born inTexas increased by approximately 1.38million. In addition, immigrants keptTexas population growing during therecent economic downturn and tepidrecovery. From April 2000 to July 2003,Texas net migration totaled 560,260,including 430,048 (77 percent) interna-tional immigrants (Table 2).

Texas is one of the most popular im-migrant gateways to the United States.

Table 1Total Population and Components of Population Change in Texas, 1950–2003

Percent change due to

Total Natural Net Percent Natural NetPopulation increase increase migration change increase migration

1950 7,711,1941960 9,579,677 1,868,483 1,754,652 113,831 24.23 93.91 6.091970 11,196,730 1,617,053 1,402,683 214,370 16.88 86.74 13.261980 14,229,191 3,032,461 1,260,794 1,771,667 27.08 41.58 58.421990 16,986,510 2,757,319 1,815,670 941,649 19.38 65.85 34.152000 20,851,820 3,865,310 1,919,281 1,946,029 22.76 49.65 50.352003* 22,103,374 1,259,945 699,685 560,260 6.04 55.53 44.47

*Through July 2003.

SOURCE:Census Bureau.

Chart 1Texas and U.S. Population Growth,1970–2003

Annual growth rate (percent)

SOURCE: Census Bureau.

United StatesTexas

0

.5

1

1.5

2

2.5

3

20031990s1980s1970s


Chart 2 shows the percentage growth ofthe foreign-born population in the UnitedStates, Texas and the state’s six major met-ros during the 1990s. The foreign-bornpopulation share in Texas rose significant-ly during the decade and in 2000 com-posed 14 percent of the population com-pared with 11 percent at the national level.

In recent years, growth of the foreign-born has been even more rapid in Texas’major metros than in its border metros.Between 1990 and 2000, the number offoreign-born in the major metros morethan doubled (112 percent increase),while that of the border metros increased51.6 percent, well below the state averageof 90.2 percent.3

Of Texas’ major metros, only El Paso(31.5 percent) and San Antonio (54.3 per-cent) recorded foreign-born growth ratesbelow the U.S. average (57.4 percent),mostly because many of the immigrantsin these metros entered the state in earlieryears and their second-generation chil-dren now reside there. Austin witnessedthe strongest growth in the foreign-bornduring the 1990s (172 percent), likely dueto the booming tech economy there. Theshare of the foreign-born in Dallas, FortWorth and Houston grew by 152 percent,131 percent and 94 percent, respectively.Shares of the foreign-born in the majormetros are shown in Chart 3.

This increase in immigration hasbrought rapid change in the state’s ethniccomposition. Because of Texas’ proximityto Mexico, many of the state’s immigrantsare of Hispanic origin. Hispanics are by farthe fastest growing segment of the popu-lation. During the 1990s, Texas’ Hispanicpopulation grew at a pace of 54 percent,adding more than 2.3 million people. As aresult, Hispanics now make up 35 percentof the state’s population, compared withroughly 14 percent at the national level.4

Among states, Texas has the country’s sec-ond-highest Hispanic population, behindonly California.

Texas’ population has changed inother ways as well. Anglos’ share of thetotal population has fallen—no longerabove 50 percent—as their rate of growthslowed in the ’90s and the first three yearsof this decade, while blacks still accountfor about 11 percent of the state’s popula-tion (Chart 4). The number of peopleincluded in the “other” category has dou-bled since the 1990s.5

Chart 2Growth of Foreign-Born, 1990–2000

Percent

0

20

40

60

80

100

120

140

160

180

San AntonioHoustonFort Worth-Arlington

El PasoDallasAustinTexasU.S.

SOURCE: Texas State Data Center.

Chart 3Share of the Foreign-Born

Percent

0

5

10

15

20

25

30

San AntonioHoustonFort Worth-Arlington

El PasoDallasAustinTexasU.S.

20001990

SOURCE: Census Bureau.

Chart 4Change in Ethnicity/Race for Texas

Percent

2003200019901980

Hispanic

Other

Anglo

Black

0

10

20

30

40

50

60

70

SOURCES: Census Bureau; Texas State Data Center.

In recent years, growth of the foreign-born has

been even more rapid in Texas’ major metros

than in its border metros.


The dramatic rise in Texas’ Hispanicpopulation (both immigrant and native)has far-reaching implications. Hispanics’higher-than-average birthrate suggeststhat this demographic segment will con-tinue to grow at a more rapid pace thanthat of Anglos and blacks, even assumingno immigration. In addition, Hispanics,on average, are younger, which has ramifi-cations for housing, education and thelabor force. In 2000, the median age ofHispanics in Texas was 25.5 versus 38 forTexas Anglos. This compares with themedian age for all Texans of 32.3 and forthe United States of 35.3. Currently,because of its Hispanic heritage, Texas isthe second youngest state in the nation,behind Utah.

Population ProjectionsTexas’ population will change in two

Chart 6U.S. Population by Age in 2003

Age group, in years

Under 5

5 to 14

15 to 24

25 to 34

35 to 44

45 to 54

55 to 64

65 to 74

75 to 84

85 and over

0 5 10 15 20 25 30 35 40 45 50Millions

Babyboomers(39 to 57years)

Echo boomers(8 to 23 years)

SOURCE: Census Bureau, American Community Survey 2003.

Chart 5Projected Proportion of Texas Population by Race/Ethnicity

Percent

2040203020202010

Hispanic OtherAnglo Black

0

10

20

30

40

50

60

NOTE: Assuming net migration rate is half that of 1990–2000.


major ways over the next several decades:in diversity and in age.

Diversity. The Texas State Data Centerprojects that by 2020, Hispanics will makeup the majority of Texas’ population,while Anglos will fall to the second-most-populous ethnicity (Chart 5). By the year2040, Hispanics will account for over 50percent of all Texans, while one-third ofthe population will be Anglo. Blacks areexpected to make up 9.5 percent of Texas’population in 2040, and other races (notAnglo, black or Hispanic) are expected togrow to almost 6 percent of the popula-tion.6

For Texas’ border cities, which alreadyhave large Hispanic populations, thechanges could be even more dramatic.For instance, El Paso, 78.2 percent His-panic now, will likely increase to 90.3 per-cent by 2040. Similarly, San Antonio, with

its ties to Mexican heritage, will movefrom 50 percent Hispanic (in 2000) to 61.1percent in 2040. Even Austin, where His-panics make up only 26 percent of thetotal today, is expected to see a majorincrease in its Hispanic population by2040—up to 44 percent.

Currently, large disparities marksocioeconomic conditions among Texas’ethnic groups. Compared with their Anglocounterparts, Texas’ Hispanics tend tohave lower levels of education, have lowerwages and depend more on state services.This is partly a result of immigration—Mexican immigrants tend to have averagewages 40 percent below those of natives.7

These wage differences reflect that theimmigrants are young, have scant jobexperience and speak little English.

While some of the difference betweenimmigrants’ and natives’ wages is madeup after substantial time in the UnitedStates, disparities between groupsremain. Without changes in socioeco-nomic conditions, this implies that Texas’future population could be less educated,less competitive, poorer and more in needof state services such as health care andwelfare. Texas’ challenge is to reduce thesesocioeconomic differences through in-creased educational attainment andtraining, so Texas can compete in thenation’s workforce in coming decades.

Age. Texas’ overall population, likethe nation’s, is growing older. This aging isa result of the maturing of the baby boomgeneration, which makes up the largestsegment of our population. In 2003, thebaby boomers spanned the ages 39 to 57(Chart 6). The youngest of the baby boom-ers will turn 60 by 2024. As they retire, thebaby boomers will put large demands onthe Social Security system and other gov-ernment programs for the elderly, such asMedicare. In addition, the boomers maydrive housing demand toward move-upor second homes as well as houses morepopular with older adults or combinedfamilies.

One factor that may mitigate Texas’aging population is that the fast-growingHispanic population has a different agestructure than the Anglo population. AsChart 7 shows, in 2000 the population inage groups over 35 was predominantlyAnglo. For example, in 2000, 66 percent ofTexans aged 55–59 were Anglo comparedwith 20 percent that were Hispanic. Con-


versely, of Texans aged 5 and under, 44percent were of Hispanic heritage, com-pared with 39 percent Anglo.

If expectations of rapid growth holdtrue for Texas’ Hispanic population, His-panics will make up a much higher per-centage of most age groups by the year2040, with only those over 65 being pre-dominantly Anglo (Chart 8).8 The age dif-ferential between the Hispanic and Anglopopulations has important implicationsfor education, housing and state services.

Demographics and Poverty Texas Becoming Poorer? Texas’ econ-

omy grew faster than the nation’s duringthe 1990s, and all sectors added jobs.Employment in Texas during this periodgrew at an annualized average rate of 3.3percent, above the nation’s 2.2 percent.Despite this phenomenal growth inemployment, Texas has the eighth highestpoverty rate in the country and has not yetachieved per capita income parity withthe nation.

During the 1990s, Texas per capitaincome grew rapidly—at an annual aver-age rate of 7.2 percent, which exceededthe nation’s 5.7 percent. Consequently,Texas, which began the decade at 89 per-cent of U.S. per capita income, edged upto 95 percent of the U.S. average by 2000.Moreover, poverty rates in the statedeclined—from 18.1 percent in 1989 to15.4 percent in 1999—thanks to a strongeconomy.

Although Texans’ incomes improvedduring the ’90s, succeeding years haveseen a reversal of this phenomenon.According to 2003 data, the Texas povertyrate rose to 16.3 percent and Texas nomi-nal per capita income fell to 93 percent($29,372) of the U.S. average ($31,632) asthe Texas economy slumped into therecession that started in 2001 and lasteduntil mid-2003. The state’s higher concen-tration of high-tech and transportationindustries, which were the hardest hit,intensified the recession’s impact. Hence,these industries shed a substantial num-ber of high-paying jobs, pushing down thestate’s per capita income more so than theU.S. average. Also, Texas’ recovery fromthe recession has been unusually weak.9

Ethnic Disparities. Among ethnicgroups, Hispanics are undoubtedly thelargest segment in poverty in Texas. In1999, more than 1.6 million (25.4 percent)

Hispanics in Texas were poor.10 Theirmedian household income was $29,873,far below the Texas average of $39,927.This is an alarming number, given theimportance of this segment to Texas’future.

Blacks had the second-highest povertyrate (23.4 percent) with a median incomeless than that of Hispanics. Anglos faredbest, with the lowest poverty rate (7.8 per-cent) and the highest median householdincome ($47,162 in 1999) in Texas.

The disparity among ethnicities whenit comes to income and poverty is not sur-prising. Natives (predominantly Anglo)are far more likely to have a high schooldiploma and some college education thanimmigrants (predominantly Hispanic).11

Less-educated individuals tend to belower-skilled workers employed in low-

paying jobs. In addition, because the non-Anglo population in Texas is far youngerthan the Anglo population, a large per-centage of non-Anglos are in their earlyearning years, have scant work experienceand thus are more likely to have lowerincomes.

ImplicationsIf the income differential between Ang-

los and non-Anglos persists, a larger shareof Texans could be drawn into poverty inthe future. According to the Texas StateData Center, the share of households withannual incomes of $25,000 or less will in-crease from 30.7 percent (in 2000) to 37.5percent by 2040. Moreover, the percentageof families with earnings exceeding $100,000will fall from 11.5 percent to 8.5 percent.The net impact could be a decline in real

Chart 7Texas Population by Age and Ethnicity, 2000

Percent

Anglo

Hispanic

0

10

20

30

40

50

60

70

65+ 60 to 6455 to 5950 to 54 45 to 4940 to 4435 to 3930 to 3425 to 2920 to 2415 to 1910 to 14 5 to 9Under 5

80

AgeSOURCE: Texas State Data Center.

Chart 8Texas Population by Age and Ethnicity, 2040

Percent

Anglo Hispanic

0

10

20

30

40

50

60

70

80

65+ 60 to 6455 to 5950 to 54 45 to 4940 to 4435 to 3930 to 3425 to 2920 to 2415 to 1910 to 14 5 to 9Under 5 Age

NOTE: Assuming net migration rate to the state is equal to that of 1990–2000.



income, reduced tax revenue per house-hold and increased burden on the stategovernment to pay for welfare services inTexas. As the state is likely to depend pro-gressively more on non-Anglo Texans forfuture tax revenues, it is important tolessen the existing wage gap and educa-tion differential between ethnic groups.

Education and the Labor Force. Oneway to reduce the wage gap is througheducation and training. In fact, accordingto the Texas comptroller, every dollar

invested in Texas’ higher education sys-tem returns $5 or more to the Texas econ-omy. Hence, it is essential that the educa-tion system keep up with the state’schanging demographics.

Texas’ education record is nothing tobrag about. Texas ranks second to lastamong the 50 states in its share of thepopulation 25 years or older with a highschool diploma (only 77.8 percent). Fur-thermore, in 2003 several Texas cities(Dallas, El Paso, Fort Worth, Houston and

San Antonio) ranked in the bottom thirdamong major U.S. cities in shares of highschool graduates.12

Again, the statistics vary by race. Forinstance, Anglos in Texas are more likelyto be high school graduates (87.2 percentin 2000) than their non-Anglo counter-parts, especially Hispanics. In 2000, morethan half the Hispanic population in Texasdid not have a high school diploma. Ang-los are also more likely to attain higherlevels of education than non-Anglos,excluding Asians. According to the PewHispanic Center, Hispanics are half aslikely as Anglos to graduate from collegewith a bachelor’s degree by age 26 (23.2percent for Hispanics versus 47.3 percentfor Anglos). Much of the disparity is dueto rapid Hispanic immigration into thestate: immigrants’ wages and educationlevels tend to be much lower thannatives’.13

Hispanics are expected to make upthe majority of the labor force in Texas by2040. If this disparity between Anglo andnon-Anglo high school and college gradu-ation rates continues, the Texas economycould face several important challenges.

First, according to the Texas StateData Center, by 2040 approximately 30.1percent of the labor force will not have ahigh school diploma, up from 18.8 per-cent in 2000.14 If that occurs, a highershare of Texas’ workforce would be lesseducated and low skilled, possibly makingthe Texas economy less competitive.

Second, empirical studies show thatlow education levels are associated withlower income levels; therefore, failure tocomplete high school or college nega-tively impacts average earnings.15 Earn-ings data from the Census Bureau demon-strate this point (Chart 9). An increasingnumber of less-educated laborers wouldreduce the average income of Texans andin turn decrease tax revenues collected bythe state.

Third, overall enrollment in publicschools is estimated to climb rapidly,growing at about half the state’s popula-tion growth rate, according to the StateData Center. Most of this increase in stu-dent enrollment—Hispanics by almost100 percent and the “other” category by 71 percent—is expected to result fromgrowth in the non-Anglo populationbecause of its younger age structure.

Thus, state expenditure on public

Where the Poor Reside in Texas

living in poverty.

In contrast, poverty levels in the major

metros have rarely been above the state

average (see table). However, they have been

higher than the U.S. average in some major

metros. For instance, since 1989, both San

Antonio and Houston have recorded poverty

rates slightly higher than the U.S. average. In

fact, Houston is home to the highest number

of poor Texans (623,493). Dallas traditionally

has posted lower poverty rates than the

nation, but the recent economic downturn

pushed its rate slightly above the U.S. aver-

age. The higher poverty rates in the Texas

border metros and some major metros may

be a result of their above-average shares of

international immigrants.

The poor live all over the state, but the

border metros fare worst, with the highest

poverty rates (see table). Although poverty

rates declined in the border metros during

the 1990s as the economy boomed, the

share of the population below poverty level

remained well above the state average of

15.4 percent in 1999. McAllen, Brownsville

and Laredo had more than 30 percent of their

population in poverty, while almost one-

fourth of those living in El Paso were poor.

The picture for the border metros has

not improved much since 1999. According to

2003 census data, Hidalgo County (McAllen

MSA), Cameron County (Brownsville MSA)

and El Paso County (El Paso MSA) rank

among the top four counties in the United

States with the highest share of individuals

Poverty Characteristics of United States, Texas and Its Major and Border Metros

Individuals below poverty Percent below povertyPlace 1989 1999 2003 1989 1999 2003

United States 31,742,864 33,899,812 35,846,289 13.1 12.4 12.7Texas 3,000,515 3,117,609 3,508,230 18.1 15.4 16.3Austin 129,942 134,589 171,373 15.9 11.1 12.8Brownsville 101,362 109,288 130,733 39.7 33.1 36.5Dallas 322,604 384,146 488,602 12.3 11.1 13.0El Paso 155,298 158,722 189,596 26.8 23.8 27.4Fort Worth/Arlington 147,177 171,930 193,427 11.0 10.3 10.7Houston 494,457 572,410 623,493 15.1 13.9 14.1Laredo 50,116 59,339 n.a. 38.2 31.2 n.a.McAllen 159,216 201,865 238,333 41.9 35.9 38.0San Antonio 252,301 234,478 266,248 19.5 15.1 16.2

NOTE: 1999 poverty data are the latest available for Laredo MSA.

SOURCES: Census Bureau; Texas State Data Center.


education as well as the number of stu-dents requiring financial assistance couldexpand rapidly unless socioeconomic dif-ferences between races are reduced. Ris-ing education costs coupled with slowgrowth in tax revenues would adverselyimpact the state’s financial situation.

However, it is naive to assume thatthe current income differential betweenAnglos and non-Anglos will persistunchecked. Empirical research shows thatsecond and third generations of immi-grants are more likely than their forefa-thers to have access to higher level educa-tion and, therefore, are better equippedwith skills required for higher paying jobs.Hence, the wage gap between non-Anglosand Anglos is likely to be reduced in thefuture.16

For the Texas economy to remainrobust, it is essential that the state’s edu-cation system make progress on at leasttwo fronts: (1) investing in resources toimprove overall student achievement,and (2) developing programs that helpbridge the educational attainment gapbetween racial and ethnic groups.

Housing. What does the future holdfor the housing industry as Texas’ popula-tion changes over the next severaldecades? The aging of the overall popula-tion, along with the baby boomers, willcertainly impact the housing industry inTexas as well as every other state. Theyoungest baby boomers turn 40 this year,and boomers are turning 50 at the rate ofseven every minute and will continue todo so through 2013 (see Chart 6). This seg-ment of the population, along with agingseniors, will be among the most potent

forces affecting the housing market andhome ownership in the coming decades.It remains to be seen what boomers’ pref-erences will be—whether they remain intheir current homes, trade up or purchasevacation homes. Most boomers are enter-ing the stage of life when earnings peak—thus, they may choose more affluenthomes or ones featuring amenities morepopular with empty nesters.

The demographic shift of the babyboom generation leaves fewer householdsheaded by those in the starter home mar-ket, ages 25 to 34, which could mean aslowdown in starter home construction.However, immigrants and minorities, whohave had historically lower home-owner-ship rates than Anglos, will likely take upsome of the slack. Home ownership isexpected to increase dramatically forminority and foreign-born households inthe coming decades, especially in areasthat have experienced high levels ofimmigration, like Texas. Because Texas’Hispanic population is younger and fastergrowing than the overall population,many Hispanic-headed households willmove into the prime home-buying agegroups in the coming decades, whichcould give Texas homebuilders a boost.

This has important implications forthe apartment market in the short run aswell, with Hispanics currently more likelyto rent than own. According to censusdata, in 2002 the U.S. home-ownershiprate for Hispanics was 48.2 percent versus71 percent for Anglos. Thus, Hispanicshave the potential to become a muchlarger segment of the home-buying mar-ket.

Chart 9U.S. Mean Earnings in 2002 by Educational Attainment

0

$20,000

$40,000

$60,000

$80,000

$100,000

$120,000

Professionaldegree

Doctoratedegree

Master’sdegree

Bachelor'sdegree

Associatedegree

Somecollege

Graduate,including GED

No highschool diploma

SOURCE: Census Bureau, Current Population Survey, 2003.

Home ownership

is expected to

increase

dramatically

for minority

and foreign-born

households in

the coming

decades, especially

in areas that

have experienced

high levels of

immigration,

like Texas.

9 Also in 2003, Texas’ median household income($40,674) was below the national average of$43,564, putting Texas 32nd in terms of medianhousehold income among the states.

10 The Census Bureau uses a threshold updatedevery year for inflation to determine the povertylevel. If an individual’s or family’s income beforetaxes and excluding capital gains or losses fallsbelow the applicable threshold, the individual orfamily is considered poor. See the CensusBureau’s web site (www.census.gov) for povertythreshold schedule.

11 See Orrenius and Viard, 2000.12 American Community Survey 2003, Census

Bureau.13 See “Immigrant Assimilation: Is the U.S. Still a

Melting Pot?” by Pia Orrenius, Federal ReserveBank of Dallas Southwest Economy, May/June2004.

14 Projection provided by Murdock et al., 2002,assuming net migration rate to the state is equalto that of 1990–2000.

15 “Educational Attainment and Border IncomePerformance,” by Thomas Fullerton, FederalReserve Bank of Dallas Economic and FinancialReview, Third Quarter 2001.

16 See Orrenius, 2004.


Texas’ housing market stands to ben-efit from its rapidly growing and diversepopulation and its strong pace of interna-tional migration. Real estate firms of thefuture will be wise to market to both theincreasingly older Anglo population andthe younger Hispanic population. Addi-tionally, while domestic migrationdropped off during the recent economicdownturn, a pickup in that segment of thepopulation would benefit Texas housing.

Health Care. The aging of the Texaspopulation plus a rapidly growing popula-tion segment with different socioeco-nomic characteristics than the previousAnglo majority will dramatically affect thehealth care industry in Texas. The numberof instances of diseases and disorders isexpected to increase in Texas. Trips to thedoctor, days in the hospital and the num-ber of people in nursing care facilities areall expected to rise at rates faster than thepopulation growth rate (Chart 10). Thehealth care industry is currently one of thefastest growing sectors of the Texas econ-omy and will likely remain so as the needincreases for long-term care facilities anddoctors who treat the elderly and a morediverse population.

OutlookDuring the 1990s, Texas grew even

faster than expected, becoming the sec-ond-largest state in the nation. Along withthis growth, the population has becomeolder and increasingly diverse, and todayit is no longer dominated by an Anglomajority. Hispanics account for thefastest growing segment of Texas’ popula-

tion and will likely make up the majorityby the year 2020. Disparities in incomeand education between Hispanics andother ethnic groups may be a challenge toTexas and its resources. The state couldreduce such socioeconomic differencesthrough increased educational attain-ment and training so that in comingdecades, the state’s workforce will con-tinue to be one of the most competitive inthe nation.

Petersen is an associate economist andAssanie is an assistant economist in theResearch Department of the FederalReserve Bank of Dallas.

NotesThe data used in this article come from two main sources, the Census Bureau and the Texas State Data Center. The two sources differ somewhat in terminology regardingrace/ethnicity. Thus, in an attempt to keepthe information consistent within the article, the authors use the terminology provided by the Texas State Data Center. For more informa-tion regarding the definitions of race/ethni-city, see http://txsdc.utsa.edu/txdata/redistrict/re-report.php and http://www.census.gov/population/www/socdemo/compraceho.html.

1 Census Bureau, International Database. For adefinition of total fertility rates, see www.census.gov/ipc/prod/wp02/appE.pdf.

2 See “The Second Great Migration: Economicand Policy Implications,” by Pia Orrenius andAlan Viard, Federal Reserve Bank of DallasSouthwest Economy, May/June, 2000.

3 Major metros exclude El Paso. The number forEl Paso has been included with the other bordermetros.

4 American Community Survey 2003, CensusBureau, www.census.gov/acs/www/index.html.

5 The term Anglos refers to non-Hispanic whitesonly. The term blacks refers to non-Hispanicblacks of African as well as non-African origin.The “Other” category includes all people who arenot Anglos, not Hispanics and not blacks. NativeAmericans, Asians and multiracial people aregrouped in this category.

6 All projections provided by “The Texas Challengein the Twenty-First Century: Implications of Pop-ulation Change for the Future of Texas,” by SteveMurdock et al., The Center for Demographic andSocioeconomic Research and Education,December 2002. Projections used in this articleassume population growth due to net migrationis half that of 1990–2000 unless specified oth-erwise. See www.txsdc.utsa.edu.

7 See Orrenius and Viard, 2000.8 Projections are based on the assumption that

the net migration rate to the state is equal to thatof 1990–2000.

Chart 10Between 2000 and 2040, Texas’ Health CareCosts Could Grow Faster Than Its Population

Percent

NOTE: Assuming net migration rate to the state is equal to that of1990–2000.


0

50

100

150

200

250

300

Nursing homeresidents

Days ofhospitalization

Physiciancontacts

Population

The Face of Texas: Jobs, People, Business,Change is published by the Federal ReserveBank of Dallas. The views expressed arethose of the authors and should not beattributed to the Federal Reserve Bank ofDallas or the Federal Reserve System.

Articles may be reprinted on the conditionthat the source is credited and a copy isprovided to the Research Department,Federal Reserve Bank of Dallas, P.O. Box655906, Dallas, TX 75265-5906. Thispublication is available on the Internet atwww.dallasfed.org.

Richard W. FisherPresident and Chief Executive Officer

Harvey RosenblumExecutive Vice President and Director of Research

Managing EditorMine Yücel, Vice PresidentResearch Department

EditorsJennifer AfflerbachKay ChampagneMonica Reeves

Design and ProductionGene AutryEllah Piña

FEDERAL RESERVE BANK OF DALLASP.O. BOX 655906

DALLAS, TX 75265–5906

jobs people business change - texas a&m university · sector’s declining importance prompt...

Documents