44

Keith M. Garison

Keith M. Cadson is an assistant vice president at the FederalReserve Bank of St Louis. Thomas A. Pollmann provided re-search assistance.

How Much Lower Can theUnemployment Rate Go?

N JUNE 1988, the civilian unemployment ratedipped to 5.3 percent, its lowest rate since May1974. The Council of Economic Advisers (CEA), inits 1988 Annual Report, forecast a continuing dropin the unemployment rate, accompanied by adecline in the inflation rate from 4.6 pci-cent in1987 to 3 percent in 1991!

These developments raise an interesting ques-tion: how low can the unemployment rate bedriven without accelerating inflation? In the late1970s, considerable research was devoted to thediscussion of such a critical rate, usually i-eferredto as the “natural rate of unemployment.” Thisresearch produced estimates of the natun-al n-ate inthe late 1970s ranging between 5 percent and 7percent, but generally were center-ed on 6 per-cent.2 With the umiemployment rate well above 6percent for most of the 1,980—87 period, the debateabout the level of the natural i-ate had subsided;with the unemployment rate moving well below 6percent in early 1988, however, the debate hasnow n-esurfaced.

This article reviews the factors that determine

the natural rate of uneniploymnent, focusing speci-fically on developnients since 1979. Fitst, it dis-cusses the concept of unemployment arid summa-

rizes how the government measur-es unemploy-ment. Second, it r-eviews the choice of benchrnar-kyears as an aid in the analysis. Finally, it examinesthe underlying determinamits of the natural rate ofunemployment in detail. Though no attempt hasbeen made to den-ive precise estimates of the natu-nal rate, the direction of its movement in recentyear’s has been detailed.

BACKGROUND: CONCEPTS ANDMEASUREMENT

To analyze recent unemployment trends, it isuseful to summarize the reasons for unemploy-ment. Since the focus is on unemployment as

measun-ed by the U.S. govei-nment, some detailabout how unemploymemit statistics ar-c gather-edis also useful tsee opposite page).

‘The CEA report was prepared in February. See Council ofEconomic Advisers (1988), p. 50. For further detail on theAdministration’s forecast, see Office of Management andBudget (1988), pp. 3b—7—8. The annual inflation rate is that forconsumer prices measured from fourth quarter to fourth quarter.

2A representative estimate is that of Cagan (1979), p. 215. For amore exhaustive survey of alternative estimates, see Weiner(1986).

3Most of the studies were done in the late 1 970s and have notbeen updated since then. See Weiner (1986). The major excep-tions are Rissman (1986) and Gordon (1987), in which he“assumes” continuation of the natural rate at 6 percent through1985. He offers statistical evidence in support of this contentionin Gordon (1988).

45

A Primer on ties. Une.nploymen t StatisticsI.,nn’Ii nnnnnnth titnn’innr~ lii’ t~i’n’kc’nni~tintniuliii’ ‘l;c’,sili,’nl as I ui, lc,sei’, ‘iur\oltlnt;ti’\ quits ‘2

thin (hR nil lIne ininrinth line L S BUreau ol lint’ Pir Ipauu’s ~oIuutan quits 3 nc-run-antsI’t)s1I’~SLnJ\(’\ s 1i5 1)1)11 inntistiiiiId’, nit’ the flu those n 110 ~t’I’~ioiisi~~urn’kt’ci. hint weri’ out ni

labor “talistit’2. Oue—tnrrrtli ol the liii’ labor Ion i’ irt’ioic’ iniokitw Ion’ t~nink ~uriclIlnjLtst’ll(riti’, I’, r’r’1,i,ir’n’cj (‘UI I) month sir lint inn on’ t’niti,inlls lutist’ ~~ini uit’~c’r~~t,n’J~c’dintuit’llnirIs(’Iicnitiisinrl(’nic’\\t’ninnnirn’u’IIrrnInIuunn’ intnI~~n’in’rur~~iooi~ini,n~tcri~~t,r’h.n]niru)tlIsjn~nrni~\iinn~iiI’nnn’n’niIni’(’aIiti\\’~Iuin’trolininnjl\ ii lint’ dali and i’t’dtnru’s hi’ ‘epoil-

I .unlol(r\’t’d in’rsonis flit’ I. tR’’so~ ~~‘lrn diii’nM lrordf’n on tannuiiu(’s. I ‘ -

rig liii’ Mil\(’\ ud’tnI~ iii itt ~~inrkid as j),iId (‘iiiIi! I’t’s~)nru1”t’Icr ,j stops iii qtit’~1iris tnt’ intel’— IIIIR1’i’s Iii hun mull hti.’,illi’ss (II Us 1lr1

1),tid

‘ ‘ui’i’ dt’tt’j’uiiirt’s n iun’ffier’ each tnt,Li’,(’inirtd ~~mo’kc’i’siii zt tatinitt —trt)i’Icnled t’iltt’n’1

ui’isu ,iiicl ‘2uinn’rruin’i’ Iii\i’Ut’s ,iuitl triuiu’r is c’iii

1nttr~i’ni iiinn’iii- lutist’ ~~innrcliii lot ~~tir’k 1)111 iu,utl julus h’nnrii

itni~t’tinil riot ill tint’ hibtit’ torte, I nt’ni1rto~n’d ~~hicin ffie~uen’e lent~nou’~n’iI~absent i,c’t’;tnnse rt

1(1 sills ,ti’e thur’,e ~~Iitilint! no t’miiIo~flut’flt illness ~~t’uitit’u’ \,tc’&itiniin shr’ikt’s on rt’t’scninal‘rug hit’ sun e\ ~~em’khint nice a~,iitainIe or reasons

lii’ jruti’n’\ J(’\tI’J’ ,ii,.nn ‘st~ttntisinu’’.~ tit’hint’i

t’,tt in nunneintpto~t’tif)I’i’si)ll I tui~ttIt’st)t’t’utitiii’ laluot’ brie N the still’ ni jni’i’’,oins aged Hi

pOints tin hod ~~trt’I~iii tine tnie’t’diiie inJiul’\~t’c’L,.- - arid o~ci vi inn dli c’k,s~ule’cI•is t’mplovt’d ui

2 vms ~~,uliut~ to Inn’ -u’’ ,tIled tir ,n oh turin- ‘ riint’iln

1uio\t’ii, 11W niiempln~nient i’att’ is catc’n—

vhn’inhi’on shn’iUirHit’t’iil,linltitIOi 3.u;N- ‘ ‘ ‘ attn as I in’ total iuLinluirei’ Itm’lnpidAe(l Lisci

v,Itinui4 icr ‘(‘1)1)11 to inn iolr ~uIinun 30 thvisrercerit (ii liii’ l;ninou’ t,ii’n’t’ L nuc’nntrnio~nun I alt’’

I tnl’tint’r iitoltlr,itiflul i, t,hui~nnt’ni,tlnircnt liii’ ai’t’ccn’tliing Iii st’s age rare and reason liii on

ie,Isn,o nil’ ornenntpinRrneini. I hi’ oot’inninlnt~it! an’ ‘liii ilos intuit cult’ ,tIstr ,‘alt’tnlzntt’ti

For a Lsok. su”ima:y of now the qo~’ern’nc’lico octs mc- qovc’rrirrc’nn a ~o provides information or lie d:snncttun-1~orrr’anonfor its unemo!oyniont ‘oport. SOC Bu”c’au of ‘ietwoei part-’nnlo aric

4u.. :‘me er”p’Oyrne’i;

Lahor Stat-sues (1987;

Unemployment can be categoi-ized as frictional,

cyclical and structur-al. Although the governmnentdoes not present its statistics in this way, such acategorization is still helpful in understandingwhy uneniployni’ient occun-s.

Cclical unemployment can be most readily

understood as representing niovements of theunemployment i’ate that iesirlt from fluctuationsof aggi-egate demand for goods and sen”u:es.Thesefluctuations, in tun-n, can he tr-aced to monetatyand fiscal policy or’ anything else that affects ag-gregate demand.

Frictional unemployment results from relativeshifts imi the supply or deniand foi goods and sei--

vices between industries or occupations. Becauseinformation about jobs is costly to obtain, peoplecami be ‘‘caught between jobs,’’ r-esultimig in tempo—

r’ary unemployment while infor-mation about othei-jobs is sought. Sometimes, to emphasize its short-i-un tn-ansitional nature, this type of unemnploy-merit is called turnover unemployment and isconsideied a vital aspect of the operation of a free-enterprise economy.

Structui-al unemployment occurs when thei-e isa mismatch of woi-ker-s and job vacancies either’ byreason of skill or location. It is only artificially dis-tinguishable flom frictional unemployment in thatit is consider-ed longer in duration and involves, inaddition to the costs of job-information seai-ch,training or relocation costs.

Categorizing unemployment into three types isa useftil way to analyze it, The three types of un-employnient involve costs in obtaining imiforma—tion about the availability of other jobs. Becauselabor- markets are chai’actem’ized by heterogeneityof skills and job r-equiremnents, it takes time and

m 0 CD = C 0 ‘C cj~

0 0 0) 0, 0 3 0 0 0) = a 0) 0 0) 0 0 0)

‘0J

CD‘0

t—

‘0o

—0

=~

~t

CD~

=—

C‘0

CP.

C

~~

c‘0 2

-~

—C

~-

0,-~

=~

CD

~P

.C(C

~-+

=-,

——

—C

z

i—

Inflationra

te

~2i

o2 3

O<

~ø

_.s

SI

-~

II

II

~I

I

aJ

/I I j

a~

95L

~~

c~~ ~

~2-5

?3

~C

Da—

:~

rC

D__C

Dt~

~~

tC

C~

P.

——

—C

C~

‘Dc—

~—

—~

2PC

~C

CD

~~2

z~—

~~

-C’~

’0~

Djt :~

.~

E~

°Z~

C’0

~C

~C

S9

9C

59

‘~

.:

P‘0

=C

~C

rDP

~4

~C

~C

D~

t’t

~~:

i~~-

~1~

C~

-.

C-~

:;~

-~

CD

~E

~—

~CD

C-D

CD

z-5

~~

~~

~zt

~t~

~’0

C~

C5

~-C

~~

~~a

—~S

t”a—

~~

r’C

~t-

~C

D~

’ 0=

—~

~‘J

_~:C

~~

Ct—

~‘C

D~

O~2

9~-

~5

•~~

t~C

D~~

~cC

5 7~

-D_•

~~

~p.

~ ~ 4p

.C~

—C

aC

t~~

=:

C-~

CC

-t-

~~

~C

:o_

?~

-C’~

- 9:C

D~

~9

~~

z~

~:

6~

z~

:p.’D

~-~

--~

-—r2

~-~

—p.~

—,~

-D

~P

~~

Or,~

”~

~—

C~

~~

~~

~ ~to

:EE

o~

9:~

9a

~~

~~

~ r—‘0

~_tr

~~

——

=~

~-~

ctp

.5p.P

.‘0

p.~

C~

~~

CD

0,_

~tp

~p.

-~—

~‘~

~-C

——

CD—

--1’

—C

CD

~;‘

ttz’9

cr~

Z-~

CE

—t~

—-C

HC

Do

~—

~~

tP.::

oC~

jCt&

r9Z

~’_

C~

~C

’~~

—~

~~_~

E_

~

~~~

~:~-

•E~

~~

C~

-;~

C~’

——

C’~

—~

P.C

~~

~o

3’0

~.a

~_

—r’D

,,C

CD

CD

r’‘-‘o

~~

CD

—~

—~

J,(

’_C

D_

.~‘D

-~5

~-

~C

’~~

C-C

~’—

—C

D~

C~

0 t~

CD

_C

C:

:~s=

~-~

;~

-~

~E

~p.

~~

5~

E~

.——

’0~

—C

.:~

r~—

C’

-tC

DC

~~

C_

2~-D

~©

~~

~:

9(~

~t-

-~tP

—’0

~( 9

’Ds_C

D:

——

4—

—~

SCD

CC-

C.

P.C -till’-

PE;-

C

E-~

~a

CJ’

C—

4—

C_

a—

)

—- —

0- S

0~

’

—0~

—

—~

‘~

o~

‘7—

0~

0 ‘a~

00 o‘~

1—

:10~

~

S CC

-c

!aur

~2

a

—‘2

—

2--,

-—C

’:

—-)

——

-~

‘.t-‘a

°

fci’

—:2

-ac

~S

~:o

a_

~ ~~—

e~r-

~C

-a—

r

o~:

IM~~

:1

tl~ r ~

i S ~t

2’

—‘

O_,c

i__i

— — ~z

C-r’

p”2-

:-c

:‘c

oC

~©

—

02

2ci,t-

’):

a~

2

5’~

a5’

C— -I

t—

~‘0

C—

—

o‘2

=2

C~

-‘—

‘c,_

t D

07

7 0H

-~

C-

f—

C-c

2

?~~. C

’2’ C

0a:

—ci

—a

-J~

‘—

-J

f’2

CC

’—

_‘c

‘t,~

ta

0C

C’

rD’2

C-

‘2~

C—

—0

-~-~

2-~

~‘:

~‘

C ~3’

2-t

,oS

~—

2:

~a

E~

~z-f

C-

C,

Oo

~

—.

—_J

—‘—

-—

-~

a-~

4-

~ S‘~

——

-~

~It

—o

—

~-

S‘2

a~-

~

——

OC

7~-

:1’

ci

2-—

‘©

0f

~“‘2

ci—

CD‘

o—

.4Z

,,-r

C’

~ —.

at—

—

~ —‘2

~‘

——

a2-

o~

‘-C

C,

0,a

-:

:~:~

=‘~

‘~

=5

r~

—C

——

‘20

C -J -i ci ci-

ci-

C— ci’

C) 0 ~0 a 0. El C)

0 0 C) Ci)

Cii

U,

0)

C C,

U,

-‘1 0 -‘I

in C (0 (0 ON

)a

~1)

—C)

ca

—.~

,r•0

—

-—

-I’

—~

0-

O_

L

03 a. C B V 0

DO

C a—

0

CD Cl)

‘U S C,

0—

Hr

ra

mor

2~tV

° o 0’

-.3

2)C

®

3 8O

aC

D 2.c

aN

a®

a0

CD5

-30’

‘<a

‘C8

°a-

0:r

cn~o

o o2

~‘2

.o m 0 c

°a O

S~

¶

PP

(00

00

0

a U,

in 0

Ca

-Ja

8 — CD~

CD

0~

CD

a‘-C

I-.’

-

o0

a ‘C CDa

Co

CDCO

(,n01

0101 a

=a

5-;:

00

12)

0)

IN 00

CM

0)

00

48

Chart 2Unemployment Rates

Annual Data

85 1990

unemployment relationship yields little informa-tion about how low the unemployment ratecan go.

Chart 2 summarizes the civilian unemploymmientrate along with that for- pr-ime-age males for the1948—87 period. Civilian unemployment in thethree benchmark year-s ranges from 4.1 per-cent ofthe labor forte in 1956 to 5.8 per-cent in 1979. Corn-pared with 1979, it appears that the economyreached full employment in late 1987 when the

unemployment rate fell below 6 per-cent; corn-pared with the ear’hier benchmnarks, however-, thereseems to be r-oom for fun-ther- employment

expansioml.

The structure of unemplovnient, especially as iti-elects a changimig composition of the labor- force,is an additional consideration in assessing thenearness of actual employment to full employ-ment. Chart a also shows unemployment forprime-age males, the group that has the lowestturnover’ r-ate in the labor’ for-ce. The unemphoy-

merit rate for this group was 2.9 percent in 1956,2.6 pertent in 1973, and a soniewhat higher 3.4per-cent imi 1979. The 1987 unemployment n-ate for’this group aver-aged 5 percent, again suggestingthere is room for- flir-ther expamision inemployment.

A dir-ect examination of those unemploymentmeasur-es that are consider-ed most iriiportant(hoes not pr-ovide a clear-—cut conclusion aboutwhether the 1987 levels of unemplovmemrt indicatean economy approaching full cniployment. ‘there-

fore, we examine the composition of the labor’force in for-ther detail.

CHANGES IN DEMOGRAPHIC

FACTORS

The overall unemployment rate r’eflects aweighted average of many umiemployment rates, astable 1 indicates. As this table shows cleam’ly, cer’-tam gr-oups typically have higher or lower ummem-

Percent Percent10 10

8

6

4

2

1945

8

6

4

50 55 60 65 70 75 80

2

Table I Table 2Unemployment Rate byAgeand Sex Composition of Labor Force by Age(selected years) and Sex (selected years)

- 1966 1973 1979 1987 1956 1979 1981

total 41% 4,9% 59 S Total 000° 000°’ 1000° 1009%

MS MS 877 610 579 6516—IS 410 139 9 178 16-19 3 6 49 3420—2 69 73 7 99 20—24 5 80 51 65

5-54 29 26 34 50 25—54 456 8 36-1 755—84 4 7 37 55—84 93 78 69 5855 3 30 4 26 65 3 21 19 6

Ferna Female 322 389 421 4416—19 110 15.3 164 159 t&—9 2 43 43 220—24 63 84 96 94 20-4 3 6 9 025—54 4 4 1 5—54 208 24 253 30455-64 7 S 3 31 5-64 3 47 4 4165 29 24 85 1 -t 11 o

Ma riot equa 00 d ato raiding

plovmnent rates than the o~em all 1\ erage. h he teen

age group is ah\avs the highcst followed by thc~ 24 ~eat old gr oup. (‘onsequenth’ to c nderstand more fully the significance of gi~enunem Tablesplovmnent rate, one must c anirm both thc m’elativeimportance of each age gm’ou~~in any gi~en year’ Growthof Labor Force by Age and Sexand the gm o’is th tate of each age group oi en time (selected years annual rates)

Iabhc 2 summarize mhc mel &tive mmportancc of 1956-’7S 197$—fl 197947the diffcr ent age-se groups for’ the benchmark ‘I’Sh 1 8° 2 7

°G I 7yeats. One striking obsen ation ms the change inthe r-atio of mahes to females that hm taken place Malesince 1956. this changrmig proportion how ter 6- 9 3 4 27rna\ m ot be cr mtir al in inter pi eting ~ hat has hap- 29

pr nod to the overall unemployment rate: as tablet 55’-M 07 04 0,5shows li mnale uneniplovment is not always ‘ibm 0 65 I 8 03 0that for malos

emma aVt hat is mrnpot tant in inter preting mno~ements in 6— 9 43 9

the unr-mplovment ratr o~ertime is the shifting 20—24 50 42 0.

imnpor tanc e of age gr-oups. Obviously the m’ise in 2544 3 48 40mnipomtance ot thc lb 19 and 20 24 ear old & 0groups horn 1956 to 1979 was an important tactor --

in intem pr eting unemplo ment tm-ends for thatperiod \s table 1 shows the unemployment rate For thc 1979 87 pet iod the 2a 54 year -old group~-~asalways highest for these groups. Significantly, grew fastest reflecting the maturation of the 16 19howc~er thcsc youngest ag gm’oups have declined and 20 24 sear old groups of thc t970s.as a pm 01)01 tion of the labor fom cc fi oni 1979 to198 Jo aid tn analyLing the effects of changes in thc

/, underlying demographics a weighted unemploytable 3 summarizes the gmowth of the labor mnent rate, where the weights are based on the

force by age group hct~een the benchmark ycars. cornposmtmon of the labor for-ce is common! used.

‘For an analysisof women in the labor mamket ee Shank See Clark (1977), Flarm (1979), Cain (1979), Antos, Mellowand(1988). Tnplett (1979) and Cagan (1979).

50

Table 4 summarizes various unemployment m-atesand provides information about how the changingcomposition of the labor for-cc influences the over-—all unemirplovnient r-ate. A comparison of the altem—native r-ates with the over’all n-ate show’s that demo—gm-aphic shifts were most pronounced in the1956—79 period.’ Changes in the coruposition ofthe lahom- for-ce slufted the unemplovmucnt taleupwam-d by 0.8 per-centage points lovem-all lessfrxecl—weight column) compan-ed with an actualrise of 1.8 percentage points 5.9 nunus 4.1). Inother words, the labor market pressure of 4.1 per-cent iii 1956 would have changed to 4.9 percent in1979 because of a shift in the composition of thelabor force toward the youngest gr-oups.

large in determining if the economy is at full erii—plovnient. For demographic reasons alone, the1987 natural rate of unemployment is only about

0.3 percentage points higher than that of 1956, and0.5 percentage points lower than that of 1979. ‘[hisinterpretation is suppor-ted by the change in thedifferential between the overall r-ate and that for-prime-age males. ‘i’hus, although methods van’ incalculating the effects of changing clemograplucson the unemployment mate, there seems to be littledoubt that changes in the composition of the la-bor- fom-ce since 1979 have producecl a lower natu-ral rate of unemployment.

Table 4

Alternative Unemployment RatesPrtme-Age OverallLess Overall Less

— OveraU Rate Fixed Weigflt Male Axed-Weight Prime-Age Male

1956 41 43 2S°o 0 °/ 12

1973 49 44 26 05 23

1979 59 53 34 0.6 251980 72 66 52 06 201981 76 71 5.5 05 2.1198 97 92 80 05 171983 96 92 82 04 141084 75 72 59 03 161985 7 70 56 02 161986 7.0 as se 02 1,41987 62 61 50 01 12

Calculated usmng veraga of labo forcecomposrtron rn 1956 and 1987 ( eatable 2)

‘there was also a widening of the diffem’ence

hetween the overall uneniplovnient n-ate and thatfor prime-age males, reaching 2.5 per-centagepoints in 1979, up frorii 1.2 percentage points in1956. This differential yields the same general con-clusion: consicler-ing demographic changes, themmtur-al n-ate of unemployment rose quite shar-plvbetween 1956 and 1979.

Since 1979, the composition of the labor’ for-cehas shifted hack toward the older groups, whictisuggests that the natut-al rate of crnernplovmerithas declined. ‘Ehe d iffer-erice between the fixed—weight measure and the over-all i-ate has nar-roweclto almost zero. The smaller- differential nieans thatclemuographic considerations no longer loom as

(~nL%!%rcE~sIN INSTI’TUTIO]NALFACTORS

As noted earlier, in addition to the age—sex com-

position of the labor force, mans’ other factorsinfluence the unemployment rate. ‘l’hese factom-sam-c clisc:ussed in this section along with a sum-mary of their recent trends.

Minimum Wage

Federal minimum wage legislation was iutt-o—ducecl hi the United States in 1938; by 1985, 80percent of the U.S. labor force was eniplovecl insectors subject to its coverage. B~’paving thelowest—irrc:oriie wom’ker’s a higher wage than the

competitive market would pa~’in the ahsemice of a

‘For a similar table and analysis, see Council of EconomicAdvisers (1978), p. 170.

IDSee Cain (1979).

51

Chart 3Minimum Wageas a Percent of Average Hourly EarningsPercent Annual Data60

55

50

45

40

35

3085 1990

minimum wage, this legislation raisecl the avem’agelevel of real wages above their competitive level.’

tn r-esponse, the quantity of labor- services sup-plied will exceed the quantity demanded, with thedifference being classified as unemployed.

When the minimum wage was first legislated in1933, it was 525 per hour andl covered about 40per-cent of the nation’s nonsupem-visor’v eruplovees.Over the 1938—SI period. it was raised 15 limes,reaching 53.35 per houm’ in 1981 and has miot beenchanged since. By 1985, 87 percent of nonsuper-vi—sory employees were subject to the mininium wage.

The mininluni wage law has had its gr-eatestefléct on teenage employment with little effect onother age gr-oups. Because the~’have fewer-skillsand less education, teenagers’ niam-ginal productsare typically below those of older-, mom-c exper-i-

enced worker-s. Consequently, the mininiuni wageis much more likely to he above the competitive

wage for this group.

‘l’o assess the impact of minimum \%-age legisla-tion, the minimum wage niust be viewed n-dative to

aver-age hourly earnings. The comparison measum-eused here is avet-age hourly earnings for workers inthe pr-nate nonagricultural sector. The movement

of the minimum wage n-dative to this rneasimm-e from1947 to 1987 is shown in chart 3.

After a lam-ge jump in 1950, the nunimnumn wagerelative to average houmiv eam’nings fluctuated be-tween 45 percent and 55 percent, befom-e droppingbelow 45 percent in 1972. It then rose froni about40 percent in 1973 to 46 percent in 1981. With thenunimuni wage constant at 5335 per’ hour since1981, however-, a steady decline in the relative

Percent60

55

50

45

40

35

301945

NOTE:50 55 60 65 70 75 80

Average hourly earnings are for the private nonagricultural sector.

“For a survey, see Brown, Gilroy and Kohen (1982).

52

Chart 4Replacement Ratio and Unemployment Rate

NOTE: Replacement ratio is the average of unemployment benefits paid weekly as a percent of average weeklyearnings in the private nonagricultural sector.

minimum wage has occurred since then, reaching

37 percent in 1987.

Although the relative ruinimum wage declined

from 1957 to 1973, the coverage incm’eased from 45percent to 75 percent, primarily because of the

rapid growth of teenager-s in the labor force. The

minimum wage may have pushed the unemploy-ment rate upward from 1973 to 1979, hut thistrend was sharply m-eversed from 1979 to 1987.

Since the last benchmark year- of1979, the nuni-

mum wage movements have had a positive effect

on the labor market; its decline has reduced the

natural rate ofunemployment.

Unemployment Benefits

An increase in unemployment benefits relative

to wages lower’s the cost of job search.” As a result,

other things equal individuals will search longerfor a job, lowering the amount of work that theyare willing to supply at a given real wage. Also,

individuals who am-c not in the labor force will be

inclined to enter it to obtain a job and he ehigihitefor unemployment benefits in the futun-e.

One important measure in assessing the effectof unemployment benefits is the ratio of averageunemployniment benefits paid weekly m-elative to

average weekly earnings.This matio is called then-eplacement ratio. Chart 4 shiows the ratio fm’om1947 to 1987 as a solid line. Because this matioshows cyclical movement throughout the penod,the unemployment r-ate is also charted hdashedline).

Gener-ally, the m-ephacement ratio and the unem-ployment rate move in tandem, From 1965 to 1973,

Percent Annual Data Percent10

42

39

36

8

6

4

33 21945 50 55 60 65 70 75 80 85 1990

“For discussion and references, see Parkin (1984) and Cagan(1979).

53

howeven’, the replacement ratio nose quite sharplyrelative to unemployment. From 1973 to 1979, itthen declined slightly relative to the unemploy-ment n-ale. But since 1984, the diven-genice betweenthese measures bias been sharp.

A closer- exanunation n-eveals that the source ofthus recent divergetice hs chiefly a slowing of aver-—age weekly earnings wbule unemployment benefitshave continued to r-ise at relatively rapid rates.Despite this development the replacement ratio

seems to have had a m-ecent upward effect on thenatur-al n-ate of unemployment. This effect isdampened somewhat by two considerations’. (1)n’ecent changes in tax law whereby unemploymentbenefits became par-tially subject to taxation in1979, and completely so in 1987, and (2) a general

tightening of eligibility m-equirements in recentyear-s” Thus, the actual value of the replacementnatio in 1987 relative to 1979 is less than shown in

the chart. It is impossible, however, to say howmuch the change is without further r-esearch.

Tayes

Anothem- factor of considerable importance indetermining unemployment trends is the role oftaxes in influencing the labor markets” Again, theanalysis is complex and the conclusions are notclear-cut. As an aid in understanding the macroec-onomic effects, it is useful to think in ten-ms of theeffects on laboi supply amid demand separately.

Focimsing fir-st on labor supply, the tax wedge is

the difference between the real wage that the em-ployer is willing to pay and the after--tax value oftbiat wage to the workers; the size of this wedge isimportant in the won-k-vs.-leisure decision thatpeople make. An increase in the tax wedge willr-educe labor services offered at a given real wageand may encourage a longer job-search by i-educ-ing the cost of being unemployed.

(In the demand side of the labor- market, ther-elevant tax is the employer’s contribution for

social insut-ance. lntn’oducing (or raising) this taxreduces the quantity of labor- demanded for agiven real wage and, because of the higher cost oflahor, may also lengthen the amount of time em-ployer-s take in searching fur won-kem-s.

Thus, increased tax rates, whether applicable toemployers or employees, reduce employment andmay incm-ease unemployment. To show what hashappened to the tax wedge, employee and em-ployer- taxes are combined into a summary mea-

sum-c and plotted against the unemployment ratein chart 5. This tax wedge measure incorpon-atespersonal income taxes (federal, state and local),

employer and employee contributions for socialinsum-ance, and sales and excise taxes’5 Using 1956as a reference point, the tax wedge has incieasedfnom about 21 pen-cent to mon-c than 32 percent by1987. The rise was relatively rapid from 1958 to1973, slightly slower fiT)m 1973 to 1979 and evenslower fi-om 1979 to 1987. These trends suggestthat taxes contributed to an increase in unemploy-ment befone 1979; since then, the tax wedge has

had little effect, except pemhaps to reduce unem-ployment somewhat since 1981.

Demand Shift

Recent reseanch has suggested that shifts inindustmy demand also have an effect on the natu-ral r-ate.” This effect is commonly called frictionalunemployment. Ifchanging tastes, technolo~’orn-dative factom- prices induce rapid shifts in indus-

try demands for- labor-, there will be greater uncer—tainty in labor markets and incmeased search timefor both the employee and the employer.

Unemployment that occurs for these reasons is

a healthy reflection of a dynamic economy. Forour discussion, however, only the long-run move-

ments in the composition of industrial output arerelevant. Chart 6 is one attempt to captume thisphenomenon; it shows the three-year movingavetage of the sum of the absolute percentage

“See Abraham (1988) and, for a state-by-state summaryof

unemployment legislation in 1987, see Runner (1988).‘~Meyer(1981).“The tax wedge for households (or suppliers of labor services) is

WW (1-t,)

P P (1+t,)

where W is the nominal wage, P is the price level. t, is thepersonal tax rate and t,is the consumption tax rate, This ex-pression can be manipulated to give

W (t,+t,)

P (1+t,)

Charl 4 shows the value of the expression following W/Pplus the employer’s contribution rate for social insurance, Forfurther discussion, see Parkin (1984), pp. 184—85.

“For a discussion and critique of this literature, see Johnson andLayard (1966). See also Lilien (1982), Lilien and Hall (1986),and Flissman (1986).

54

Chart 5Estimate of Tax Wedge and Unemployment Rate

30

25

20

151945

change in sector-al employnuenmt sham-es,” By thismeasur-e, them-c was a downward trend in the de-gm-ce of shifting enuploymrment until the nud-1960s;since tluen the measinre of shifting employmenthas nuoved upwamd, although it has varied sub-stantially around the trend.

Focusing on the benchmark years, there is aslight downward movement from 1956 to 1973 to1979, followed by an upward movement from 1979to 1987. From 1982 to 1987, howevem-, the measuredropped sharply.” Viewed in this perspective, it is

imnhikel that shifts in the structum’e of the economyhave influenced the natural rate of unemploymentsubstantially. The n-elationship between “demand”shifts and thue unenuployment rate appean-s, ratbuer,to be a shorten--i-un phenomenon.

Other Factors

Tbue above list of factors, while not exhaustive,

sumnman-izes most of the factor-s that influenceunemployment trends. Government regulations,however, also affect labot mar-kets. For- example,

“The measure of demand shift (in year t) is

ioo~[! ~t--~H1-where E,, is employment in the ith industry in year t and E, istotal employment in year t. Data used were employees onnonagricultural payrolls by major industry. See Council ofEconomic Advisers (1988), pp. 296—97.

“One factoroperating during this period was the sharp swing inthe value of the dollar, rising sharply from 1980 to 1985, andthen falling sharply to 1987. If these developments affected

mainly manufacturing exports, the effect could have been toraise frictional unemployment (and the natural rate) when thedollar was rising, and lower such unemployment when it wasfalling. But since the focus here is on 1987 vs. 1979, a periodover which the trade-weighted exchange rate rose only 10percent, there would seem to be little net effect on the naturalrate. Furthermore, manufacturing employment as a percent oftotal non-agricultural employment has shown little sensitivity toexchange rate movements, even of the magnitude experiencedin the 1980s,

Percent Annual Data35

Percent10

8

6

4

250 55 60 65 70 75 80 85 1990

55

Chart 6Measure of Demand Shift and Unemployment Rate

8

6

4

275 80 85 1990

n-egulations imposed by the Occupational Safety

and Health Administr-ation OSFh~~)in the interestof safety and healthu can diver-t funds that wouldnuom-mnallv he used for’ investment spending.” Theser-egulations can act like an employer- tax, driving a

wedge between the wage the employer’ is willingto pay and the actual cost,

Another- example of regulations that had animportant effect on unemployment are work regis-tnation requirements for- various government pro-grams like welfare and food stamps. For- example,in 1972, legislation was passed that n-equired wel-fare mothers who were able to work to r-egister forwork,” Althuough some found jobs, others wemeadded to the count of the unemployed.

SUM%’EARY OF FACTORS AFFECTINGNATURAL RATE

The role of these factors is bmouglit together’ in

tables. Shown am-c genen-al conclusions about the

directionu in which the stn-uctural facton-s have beenoperating between the benchmar-k year-s simuce1956, No attempt is made to estimate precise1~’tbuenuagnitude of the effects on the natimr’al rate ofunenuployment.

The nuost obvious change in recent years is theshifting age distribution of the labom’ force, whichhas reduced the unemployment i-ate. In other-words, the baby-boomer-s, who nuade their- pres-ence felt thm-oughout the 1970s by pushing up thuenalur-al n-ate of unemployment, are now in thepn-ime-age gm-oup. Having accumuuulated skills, edu-cation and expem-ience, this gn-onmp is now man-ket-ing its productive skills, thus reducing the natun-al

rate of unemuployment by about one-half of a per-cenutage point fr-om 1979.

The minimum wage has had a favor-able effect inreducing the trend of unemploynnent since 1981,but no attempt was made huer-e to estinnate themagnuitude of effect. Cagan, however, estinuated

Percent10

1945 50 55 60 65 70

“For a broad survey, see Licht (1988). “Clarkson and Meiners (1977).

56

\‘1\\\\\t$&~t3 ~t9~79’~4\

World %‘%‘am’ It period. In gener-al, the key factor-sthat i njluence the nat ur’aF m-ate of u muernploynueruthave served to reduce it in the 1 980s. As a result,the cur-m-emrt natural n’ate appears below the 6 per—cenut i-ate estimated in 1979.22 Shifts in the agestr-uctur-e of the Labor fom-ce alone have m-educed itabout one—halfof a percentage point. Other favor-—

able developments, as muoted in tahule 5, mnav havereduced it even fn.rrtluem’.tethgahhqSft ~, 1,

~ \~\

~\~\1\\ 1&~”~ ~tr ~‘

1sw0g9e2\ ~\\ ~, \\I\\\\~\I:A,\

/ Demattlish ~ - ~ ~I\-p 1 ,~ ,. L -~ ,‘,‘ ~‘ /

NOTE Arrnw rno’cates erect-on of efteet on the nat,na’ ‘atecm ur’errroloymc’rt Questron -‘ arc w.rn arrow nor-cates tie effec! rs probably small A zero -idr~atesnoeffect on the ,latur-a rate

that the nuinimum wage coirrr-ibuted to an in-crease in the natum-al rate of .45 percentage pointsfrom 1956 to 1977.2 The relative minimum wage in1987 was below that in 1956-With the decline inthe propom-tion of teenager-s in the lahon- force,however, the magnitude of the effect on the natu-n-al rate of unemployment is probably less thanCagan estimated.

Unemployment benefits generally- appear- tohave affected n-ecent unemployment tm-ends nega-tively. The n-eplacement ratio has risen quite dra-matically since 1984. This is misleading, however-,because starting in 1987, unemployment benefits

became hilly taxable by the federal gover-nment,while eligibility requirements have been tightenedin recent years. These developments have raisedthe cost of being unemployed and have n-educedthe trend of unemployment.

‘Faxes wer-e a factor’ in the 1956—73 period (andto some extent from 1973 to 1979), increasing un-emnpioyment, both by reducing the cost of the jobsearch (reducing fon-egone eamnings) and increas-ing the tax wedge between what employers pay to

labor- and wor-kers receive- Since 1979, however,the upward trend of taxes has slowed, suggestingthat the tax wedge has not worsened. These devel-opments are assessed as having no effect on thenatun-ah rate in the 1979—87 peniod.

Despite considerable fluctuation in the shares ofsector employment, demand-shift factors do notappear to have been a factor during the post-

L’nenuployment r-ates below 6 pem-cenut iii late1987 and ean’ly 1988 have raised questiomus abouthow far- thue m-ate can fall before inflation againemem-ges. The fact that inflation has shown noclcam’ signs of acceleration suggests that structum-alchuanges in the U.S. economy have n-educed thenatural rate of unemnployment below what it wasin 1979.This am-tide examined sonuue of these struc—

tural facton-s.

Sever-al of these factors were found to have n-e-duced the natural rate of unemployment in recent

years, when comnpan-ed withu pn-evious experiencefrom 1956 to 1979. ‘i’he age composition of thelabor force, the minimum wage, individual andenuphoyer tax rates am-c a few of the factors thathave moved favorably. Any conclusions about un-employment benefit ratios, however, require fur--them- study, For the unemployment rate to con-tinue to decline depends crntically on the courseof future government actions, namely, legislationn-elating to the nuinimum wage, tax nates and urn-employment benefits.

-L

~4mme!2oftt1eets*nthe14atratOt+a ~

REFERENCES

Abraham, Katharine G. “Has the Natural Rate of unemploy-ment Fallen?” mimeo (University of Maryland, June 1988).

Antos, Joseph, Wesley Mellow, and Jack E. Triplett. “What is aCurrent Equivalent to Unemployment Rates of the Past?”Monthly Labor Review (March 1979), pp. 36—46.

Brown, Charles, Curtis Gilroy and Andrew Kohen, “The Effectof the Minimum Wage on Employment and Unemployment,”Journal of Economic Literature (June 1982), pp. 487—528.

Bureau of Labor Statistics, How the Government MeasuresUnemployment, Report 742 (U.S. Government Printing Office,September 1987).

Cagan, Philtip. Persistent Inflation: Historical and Policy Essays(Columbia University Press, 1979).

Cain, Glen G. “The Unemployment Rate as an EconomicIndicator,” Monthly Labor Review (March 1979), pp. 24—35.

“Cagan (1979). “Weiner (1986) and Cagan (1979).

57

Clark, Peter K. “A New Estimate of Potential GNP,” The 1977Economic Report of the President, Hearings before the JointEconomic Committee, Congress of the United States, Part 1(GPO, 1977), pp. 39—55.

Clarkson, Kenneth W.. and Roger E. Meiners. Inflated Unem-ployment Statistics: The Effects of Welfare Work RegistrationRequirements (University of Miami School of Law, March1977).

Council of Economic Advisers, Economic Report of the Presi-dent (GPO, 1978).

________ Economic Report of the President (GPO, 1988).

de Leeuw, Frank, and Thomas M. Holloway. “Cyclical Ad just-ment of the Federal Budget and Federal Debt,” Survey ofCurrent Business (December 1983), pp. 25—40.

Ftaim, Paul 0. “The Effect of Demographic Changes on theNation’s Unemployment Rate,” Monthly Labor Review (March1979), pp. 13—23.

Gordon, Robert J. Macroeconomics. 4th ed. (Little, Brown andCo., 1987).

________ “The Role of Wages in the Inflation Process,”AmericanEconomic Review, Papers and Proceedings (May1988), pp. 276—83.

Johnson, G. E., and P. R. G. Layard. “The Natural Rate ofUnemployment: Explanation and Policy,” in Orley Ashenfelterand Richard Layard, eds., Handbook of Labor Economics, vol.II (North-Holland, 1986), pp. 921—99.

Lilien, David M. “Sectoral Shifts and Cyclical Unemployment,”Journal of Political Economy (August 1982), pp. 777—93.

Lilien, David M., and Robert E. Hall. “Cyclical Fluctuations inthe Labor Market,” in Orley Ashenfelter and Richard Layard,eds., Handbook of Labor Economics, vol.11 (North-Holland,1986), pp. 1001—35,

Meyer, Laurence H., ed, The Supply-Side Effects ofEconomicPolicy, Proceedings of a Conference cosponsored by Wash-ington University in St. Louis and Federal Reserve Bank of St.Louis, October24—25, 1980 (May 1981).

Office of Management and Budget. Budget of the United StatesGovernment: Fiscal Year 1989 (GPO 1988).

Parkin, Michael. Macroeconomics (Prentice-Hall, 1984).

Rissman, Ellen R. “What Is the Natural Rate of Unemploy-ment?” Federal Reserve Bank of Chicago Economic Perspec-tives (September/October 1986), pp. 3—i 7.

Runner, Diana, “Changes in Unemployment Insurance Legis-lation During 1987,” Monthly Labor Review (March 1988), pp.9—16.

Santomero, Anthony M., and John J. Seater, “The Inflation-Unemployment Trade-Off: A Critique of the Literature,” Jour-nal of Economic Literature (June 1978), pp. 499—544.

Shank, Susan E. “Women and the Labor Market: The LinkGrows Stronger,” Monthly Labor Review (March 1988),pp. 3—8.

Weiner, Stuart E. “The Natural Rate of Unemployment: Con-cepts and Issues,” Federal Reserve Bank of Kansas CityEconomic Review (January 1986), pp. 11—24.

Licht, Walter. “How the Workplace has Changed in 75 Years,”Monthly Labor Review (February 1988), pp. 19—25.