supplemental measures of material well- being: basic needs ...consumer durables from 1992 to 2002,...
TRANSCRIPT
INTRODUCTION
This report continues an effort begun in2003 to explore additional ways ofassessing the economic well-being of theU.S. population. Issued as a companionto the U.S. Census Bureau’s reports onalternative measures of income andpoverty, this report also attempts toexpand our understanding of the popula-tion in poverty in the United States.
The economic well-being of a populationcan be measured in many ways. Usingincome to measure well-being avoidscomparing the situations of people withdifferent spending preferences. Forexample, one person might spendmoney on a variety of goods; anothermight spend almost all on video games;while a third might spend little and savefor a rainy day. If their incomes are thesame, an income measure considersthem to be equally “well-off,” even if theydiffer in other aspects of their materialcondition.
Poverty, as officially defined, indicatesthat a family’s income is below a thresh-old designed to reflect its needs. Thethresholds, designed in the 1960s, varyby family size and composition to deter-mine who is in poverty. If a family’s totalincome is less than that family’s thresh-old, then that family, and every individualin it, is considered to be in poverty.Official poverty estimates are based ondata collected in the Current PopulationSurvey’s (CPS) Annual Social and EconomicSupplement (ASEC), following the Office of
Management and Budget’s (OMB)Statistical Policy Directive 14 (1978).
Income-based measures do not alwaysprovide a completely satisfactory meas-ure of material well-being. Cash incomemeasures, which are used in mostCensus Bureau reports, do not includethe effect of taxes and noncash benefits,nor do they account for the advantagesof owning long-lived assets such as ahome or consumer durables.
Another measure of economic well-beingis expenditures of households and fami-lies.1 Expenditure measures focus onwhat a household spends rather than onincome it receives. Closely related are“consumption” measures, which focus onthe benefits derived from money spent oritems owned. A report on poverty meas-urement by the National Academy ofSciences stated that expenditures andconsumption measure not only the abilityto maintain a certain level of well-beingbut also the actual level attained (Citroand Michael, 1995). Researchers haveobserved that expenditures vary less thanincome in the United States, because peo-ple tend to save when their income ishigh and spend from savings or borrowwhen it is low (Slesnick, 1993). Thus,information on expenditures can shedlight on economic well-being beyond thatwhich comes from information on currentincome alone.
U S C E N S U S B U R E A UHelping You Make Informed Decisions
U.S. Department of CommerceEconomics and Statistics Administration
U.S. CENSUS BUREAU
Issued December 2005
P23-202
Supplemental Measures of Material Well-Being: Basic Needs, Consumer Durables,Energy, and Poverty, 1981 to 2002
CurrentPopulationReports
Special Studies
1 See the discussion in the 2003 report on sup-plemental measures of well-being (U.S. CensusBureau, 2003).
The focus of this report is on yetanother set of measures of eco-nomic well-being, simply called“material well-being” indicators,which provide a more or less directpicture of the conditions in whichpeople are living. The indicatorsinclude the presence or absence ofselected appliances and electronicgoods in a home, take account ofhousing and neighborhood condi-tions, obtain ratings of communityservices, and measure whetherhouseholds have been able to meetbasic needs. Some specific exam-ples include items such as refriger-ators, telephones, and air condi-tioning; problems such as brokenwindows, leaky roofs, fear ofcrime, streets out of repair; andwhether rent and utility bills havebeen paid. Overall, the set ofmeasures is meant to take broadaccount of a household’s standardof living. Material well-being meas-ures have some features that makethem attractive as an additionalmeasure of economic well-being.For example:
■ Material well-being measurescan be closer to conditionsexperienced at the present time.More like expenditure measures,and less like income andpoverty measures, they are lesssubject to major swings whenpeople temporarily have high orlow income.
■ Material well-being can beimproved by purchases made inprevious years, such as appli-ances, cars, or houses. Peoplebenefit from these assets longafter they received the incomeor made the purchase. Materialwell-being measures are closerto consumption measures thanare income, poverty, or expendi-ture measures in reflecting thisaspect of living conditions.
■ Some measures of material well-being reflect not only income orother resources, but also needs,which are sometimes hard tomeasure. For example, peoplemay have health problems, ahigh cost of living, or other fac-tors that make it harder to makeends meet with a given level ofincome. When needs put a strainon the budget, the strain affectsliving conditions and thereforeaffects material well-being.
■ Material well-being can also differfor households depending onassistance provided by relatives,neighbors, employers, charities,and others not counted inincome or expenditures.
The 2003 report SupplementalMeasures of Material Well-Being:Expenditures, Consumption, andPoverty 1998 and 2001 was issuedby the Census Bureau with the col-laboration of the InteragencyWorking Group on AlternativeMeasures of Material Well-Being,which included representatives ofthe Office of Management andBudget, the Bureau of EconomicAnalysis, the Bureau of LaborStatistics, and the EnergyInformation Administration (U.S.Census Bureau, 2003). That reportprovides further background anddiscussion of detailed properties ofwell-being measures and a bibliog-raphy of related materials.
This report focuses on trends ineconomic well-being from 1992 to2002. The purpose is to broadendiscussion of the issues involvedwith supplementing income-basedpoverty measures with other meas-ures that focus on consumption andmaterial well-being. It is by nomeans a comprehensive document.
The next section discusses trendsin material well-being from 1992 to
1998, using data from the Surveyof Income and ProgramParticipation (SIPP), and includesinformation on material well-being,the relation of material well-beingto poverty, and changes for popu-lation subgroups. The third sec-tion examines the possession ofconsumer durables from 1992 to2002, using data from theConsumer Expenditure Survey (CE).The fourth section uses data fromthe Residential EnergyConsumption Survey (RECS) toexamine trends in energy use andenergy efficiency in the UnitedStates from 1981 to 2001, with afocus on energy use by householdswith poverty-level incomes. Thereport concludes with a review ofresults, a discussion of datasources, and information on howto obtain additional detailed tablesand other information related tothe subjects covered in this report.
TRENDS IN MATERIALWELL-BEING: 1992 TO 1998
The Survey of Income and ProgramParticipation (SIPP) provides meas-ures of more than 70 items relatedto household material well-being.Thirty-nine of the measures reportedin 1992 and 1998 can be comparedand are examined here.2 Thesemeasures cover five topical areas or“domains”: consumer durables,
2 U.S. Census Bureau
2 The data for this section of the reportcome from the 1991, 1992, and 1996 panelsof the SIPP. Data from the 1991 and 1992panels used in this report were collected fromOctober 1992 to January 1993; data from the1996 panel used in this report were collectedfrom August to November 1998. The popula-tion represented (the population universe) isthe civilian noninstitutionalized population liv-ing in the United States. The estimates in thisreport are based on responses from a sampleof the population. As with all surveys, esti-mates may vary from the actual valuesbecause of sampling variation or other fac-tors. All comparisons using SIPP data in thisreport have undergone statistical testing andare significant at the 90-percent confidencelevel unless otherwise noted.
housing conditions, crime andsafety, neighborhood conditions,and meeting basic needs. Across 28of these 39 measures, positive indi-cations of material well-being werein evidence for a higher proportionof U.S. households in 1998 than sixyears earlier (Table 1).
The trend towards greater materialwell-being was especially strong foritems that could be consideredinnovations. Bresnahan and Gordon(1997, page 2) noted that“Innovations are important if theymake a difference in the wayhuman beings live and work . . .
New goods matter if they improvethe quality of life . . . and/or allowthe current quality of life to bemaintained at less expense.” Thedurable goods considered innova-tions are computers, microwaveovens, and videocassette recorders(VCRs). In 1992, 21 percent of
U.S. Census Bureau 3
Table 1.Percent of Households With Selected Indicators of Material Well-Being: 1992 and 1998
Indicator
1992 1998
Percentage-point changeEstimate
90-percentconfidence interval Estimate
90-percentconfidence interval
Consumer durablesFood freezer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37.1 36.5−37.7 34.9 34.2−35.7 *–2.2Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.7 20.3−21.2 42.0 41.4−42.6 *21.3Dishwasher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49.3 48.6−50.0 56.0 55.4−56.7 *6.7Air conditioner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68.7 68.1−69.5 77.7 77.0−78.4 *9.0Clothes dryer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78.0 77.5−78.4 77.8 77.3−78.3 –0.2Clothes washer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84.8 84.4−85.2 82.0 81.7−82.4 *–2.8Videocassette recorder . . . . . . . . . . . . . . . . . . . . . . . . . . 73.8 73.4−74.3 85.2 84.7−85.7 *11.4Microwave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.2 81.7−82.6 90.7 90.3−91.1 *8.5Telephone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94.7 94.5−94.9 96.2 95.9−96.4 *1.5Television . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96.6 96.4−96.8 98.4 98.2−98.5 *1.8Stove. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98.9 98.8−99.1 98.7 98.6−98.8 –0.3Refrigerator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99.1 99.0−99.2 99.3 99.2−99.4 0.2
Housing conditionsNo problem with mice, rats, insects . . . . . . . . . . . . . . . 85.1 84.7−85.6 87.3 86.8−87.8 *2.2Housing not so bad, would move . . . . . . . . . . . . . . . . . 90.0 89.7−90.3 92.1 91.8−92.5 *2.1No roof or ceiling leaks . . . . . . . . . . . . . . . . . . . . . . . . . . 91.4 91.1−91.7 93.1 92.8−93.4 *1.7No broken windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.4 92.1−92.7 95.9 95.7−96.1 *3.6No cracks in walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95.4 95.1−95.6 96.0 95.8−96.2 *0.7No plumbing problems . . . . . . . . . . . . . . . . . . . . . . . . . . 95.0 94.7−95.3 97.4 97.2−97.6 *2.4No holes in floor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98.9 98.8−99.0 99.1 99.0−99.2 0.2No exposed wires. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98.6 98.4−98.7 99.2 99.1−99.3 *0.6
Crime and safetyDoes not stay at home for safety . . . . . . . . . . . . . . . . . 89.2 88.9−89.6 87.1 86.8−87.5 *–2.1Neighborhood is considered safe . . . . . . . . . . . . . . . . . 91.0 90.7−91.4 91.4 91.1−91.7 0.3Does not carry anything to protect self . . . . . . . . . . . . 88.5 88.2−88.8 92.5 92.1−92.8 *4.0Home is considered safe . . . . . . . . . . . . . . . . . . . . . . . . 94.1 93.8−94.4 95.9 95.7−96.2 *1.8
Neighborhood conditionsFree from traffic noise problems . . . . . . . . . . . . . . . . . . 75.4 75.0−76.0 78.6 78.1−79.2 *3.2Free from street repair problems. . . . . . . . . . . . . . . . . . 80.2 79.7−80.8 83.6 83.1−84.1 *3.4Free from trash or litter in area . . . . . . . . . . . . . . . . . . . 88.8 88.4−89.1 91.8 91.5−92.1 *3.1No abandoned buildings in neighborhood . . . . . . . . . . 89.6 89.3−90.0 92.0 91.7−92.4 *2.4No problem industry or business. . . . . . . . . . . . . . . . . . 90.4 90.0−90.8 92.7 92.4−93.1 *2.4No smoke or odors in neighborhood . . . . . . . . . . . . . . 92.6 92.4−93.0 95.1 94.8−95.4 *2.4Would not move due to poor community services . . . 94.3 94.0−94.5 98.2 98.0−98.3 *3.9
Meeting basic needsNo unmet essential expenses . . . . . . . . . . . . . . . . . . . . 85.9 85.4−86.2 86.0 85.6−86.4 0.1Expect help from friends, family, or community . . . . . 84.4 83.9−84.7 86.5 86.1−87.0 *2.1No unpaid utility bills . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89.9 89.6−90.3 90.9 90.5−91.2 *0.9No unmet need for dentist . . . . . . . . . . . . . . . . . . . . . . . 90.3 89.8−90.5 92.1 91.7−92.5 *1.8No unpaid rent or mortgage . . . . . . . . . . . . . . . . . . . . . . 92.2 91.9−92.5 94.6 94.4−94.9 *2.4No unmet need for doctor. . . . . . . . . . . . . . . . . . . . . . . . 92.3 91.9−92.6 93.9 93.6−94.2 *1.6Phone was not disconnected . . . . . . . . . . . . . . . . . . . . . 96.4 96.2−96.6 96.2 95.9−96.4 –0.3Had enough of food wanted. . . . . . . . . . . . . . . . . . . . . . 97.6 97.5−97.8 97.8 97.6−98.0 0.2
* Percentage-point change from 1992 to 1998 is statistically significant from 0 at the 90-percent confidence level.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
households had computers; by1998, 42 percent did. VCRs andmicrowave ovens made their com-mercial debuts in the 1970s, whichmakes them newer than most otheritems asked about in the SIPP.3
They increased in prevalence by 11percentage points and 9 percentagepoints, respectively, in the 1992 to1998 period. Air conditioning(though not a new technology) alsoincreased in prevalence during thistime.4
Some indicators of material well-being, however, showed no growthor even a decline. Fewer house-holds had freezers in 1998 than in1992. A greater number of house-hold respondents in 1998 than in1992 said they sometimes stayedat home out of concern for their safety.5
Many of the items that showedslow growth or no growth between1992 and 1998 had alreadybecome nearly universal by 1992.This was true of televisions,stoves, refrigerators, and freedomfrom housing problems such ascracks in walls, holes in the floor,and exposed wires. It was alsotrue of avoiding disconnection oftelephone service and havingenough food in the household. Atleast 95 percent of householdsalready exhibited these indicatorsof material well-being in 1992.6
Neither the slow growth nor the fastgrowth items clustered in any one ofthe five domains. In fact, the con-sumer durables domain included theitem with the fastest growth (com-puters) and several of the items withthe greatest declines.7 One domainshowed relatively equal growth
across items: all neighborhood con-ditions showed growth of 2 to 4 per-centage points.
While the list of 39 measured itemsprovides a detailed accounting ofchanges in material well-being,focusing on a few items helps illu-minate the change. The 2003report on supplemental measuresof material well-being illustratedthe range of available indicatorsby choosing 14 measures (U.S.Census Bureau, 2003, p. 9). Ofthese, 13 can be tracked from1992 to 1998. As with the full set
4 U.S. Census Bureau
3 Microwave ovens were introduced in1967, but less than 1 percent of householdspossessed one until after 1971 (Liegey,2001). VCRs were introduced in the late1970s (Liegey and Shepler, 1999).
4 The increase in the possession of airconditioners was not statistically differentfrom that of VCRs or microwave ovens.
5 The decline in the possession offreezers was not statistically different fromthe decline in not staying home for safetyreasons.
6 Although the percentage increase wassmall, growth was statistically significant inthe percentage of households possessing tel-evisions, having no cracks in walls, or nothaving exposed wires.
Figure 1.Households by Specific Measures of Material Well-Being: 1992 and 1998
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
Computer
Dishwasher
Air conditioner
Clothes washer
No unmet essential expenses
No abandoned buildingsin neighborhood
No roof or ceiling leaks
No unmet need for doctor
No unpaid rent or mortgage
No smoke or odorsin neighborhood
Telephone
Had enough of food wanted
Stove
19921998(In percent)
20.7
42.0
49.3
56.0
68.7
77.7
84.8
82.0
85.9
86.0
89.6
92.0
91.4
93.1
92.3
93.9
92.2
94.6
92.6
95.1
94.7
96.2
97.6
97.8
98.9
98.7
7 The apparent decline in the possessionof clothes washers and dryers may have beenproduced by a change in wording of the ques-tions between 1992 and 1998. Freezers werenot statistically different in their rate ofdecline from clothes washers and dryers, andalso not statistically different from someitems in other domains: staying at home forsafety, meeting essential expenses, or havingphone service disconnected.
of items, the proportion of thepopulation that possessed itemsin most of these areas increased(Figure 1). This is especially evi-dent in the three items that werepresent in the smallest portion ofhouseholds: computers, dishwash-ers, and air conditioners (centraland room). Noticeably fewerhouseholds possessed these itemsin 1992 than in 1998. Overall, theproportion of households withthese measures of material well-being increased for 10 of the 13types.8
Evaluating changes in material well-being
This record of increasing materialwell-being measured in the SIPPmatches up with Census Bureaureports of trends in income andpoverty measured in the CurrentPopulation Survey (CPS) over thisperiod of time. While many types ofmaterial well-being were growing inprevalence, household income wasrising and poverty was falling.Median household income meas-ured in the CPS in 1992, adjusted to1998 dollars, was $38,482, and itgrew to $42,844 in 1998 (DeNavas-Walt et al., 2003). The officialpoverty rate, also based on CPSdata, was 14.8 percent in 1992 andfell to 12.7 percent in 1998 (Proctorand Dalaker, 2003). It should benoted that differences in datasetslimit the accuracy of comparisons oftrends in material well-being withtrends in income and poverty.9
Nonetheless, the increases in thesemeasures of well-being from 1992to 1998 did correspond with similartrends in other measures of eco-nomic well-being. The agreementof all three indicators increases con-fidence that overall well-being didincrease during this period.
That American households hadhigher income, a lower prevalenceof poverty, and increasing pres-ence of many indicators of materialwell-being suggests that Americanswere, in fact, “better off” in a largersense. However, various sub-groups of the population may haveparticipated in these trends differ-ently—a possibility that will beexamined below.
For durable goods, and perhaps cer-tain other material well-being meas-ures, improvements in quality andtechnology pose another problemfor interpretation. This is becausethe measures used here reflect onlywhether a household owns adurable; they do not provide anyinformation about its characteristicsor condition. Changes in materialwell-being may be understatedbecause quality improvements arenot counted, as is often the casewhen the possession of appliances,vehicles, and computers isrecorded. Alternatively, low-incomehouseholds may possess items ofolder vintage that are more expen-sive to use. One indicator relatedto the latter hypothesis, the energyefficiency of home heating and cool-ing, is examined in Section Four,which presents statistics from theEnergy Information Administration(EIA).
Finally, indicators of material well-being may partially reflect tastes orpriorities rather than economic con-ditions. In some cases, for exam-ple, a household may not want anitem and may even consider itselfbetter off by not owning it. More
broadly, changes in the proportionof households owning various itemsmay reflect general popularity andnot simply a change in ability toafford them. It may be that popu-larity or changing tastes influencedthe observed decrease in ownershipof food freezers as well as the rapidincrease in ownership of computers.Other observed patterns of materialwell-being indicators may also havebeen subject to similar influences.
Material well-being and poverty
Just as measures of material well-being increased from 1992 to 1998for all households, they alsoincreased for households withincome below the poverty linemeasured in the SIPP data10
(Table 2). Households below thepoverty line had large increases inthe possession of computers andair conditioners (the difference inthe percentage-point growth inthese two items was not statisticallysignificant). The percentage ofhouseholds below the poverty linethat did not pay rent or mortgagebecause of difficulty meetingexpenses decreased by 5 percent-age points from 1992 to 1998; thepercentage with telephonesincreased by 6 percentage points.Of the 12 measures examined inTable 2, 8 showed a significantincrease in the 1992–1998 period.
Household characteristics andmaterial well-being
Levels and trends in material well-being differ by age, sex, race,
U.S. Census Bureau 5
8 The exceptions were unmet essentialexpenses, having enough food, and posses-sion of a stove. None of these measuresdecreased significantly from 1992 to 1998.
9 Household poverty trends recorded inthe SIPP did not match trends in officialpoverty rates. Official poverty rates fell dur-ing this period, but the SIPP householdpoverty rates remained unchanged. Two fac-tors may be responsible for this difference.First, poverty calculated on a household basishas quite different properties from the officialrates, which are calculated using the family asthe unit of analysis (Mayer and Jencks, 1993;Bauman, 1999). Second, numerous differ-ences in data-collection methods create differ-ences in estimated income and other charac-teristics between SIPP and other surveys(Coder and Scoon-Rogers, 1996).
10 In this section, poverty refers to house-hold poverty as measured in the SIPP, whichis similar to the official poverty measuremeasured in the CPS, but is based on theage and income of all household members,rather than only those related by birth ormarriage. See discussion in footnote 9.
For a technical discussion of ways tocompare changes in the material well-beingof poor and non-poor households, seeAppendix C.
Hispanic origin, and other factors.11
To examine these differences, fourindicators of well-being wereselected to summarize trends intheir respective domains.12 The firstindicator measures whether ahousehold has all of the followingappliances: a telephone, a stove, arefrigerator, a clothes washer, aclothes dryer, and a dishwasher. In1992, 45 percent of all householdshad a complete set of these basicappliances. By 1998, 50 percentdid (Table 3). The second indicatorwas whether a household lived in a
dwelling free from broken windows,a leaky roof, cracks in the walls,holes in the floor, plumbing prob-lems, exposed wires, and problemswith pests. In 1992, 73 percent ofhouseholds reported they had noneof these problems, growing to 79percent in 1998. The third indica-tor, representing the domain ofcrime and safety, was whether thehousehold considered its neighbor-hood to be safe. This measureshowed no change from 1992 to1998, remaining at 91 percent inboth years. Finally, the fourth indi-cator, having fewer than two diffi-culties meeting basic needs, grewfrom 88 percent to 90 percent inthe 1992 to 1998 period.
Different segments of the popula-tion had different levels of theseindicators of material well-being.Among those with higher levels in1998 were households with house-holders aged 65 and over, whichwere more likely than youngergroups to be free of housing prob-lems and to report fewer than twodifficulties meeting basic needs(although they were below youngergroups in their rate of possessionof a complete set of appliances).13
Others with high levels of thesematerial well-being indicators in1998 were households with house-holders holding bachelor’s orhigher degrees and married house-holds without children.14 Groupswith low material well-being indica-tor levels included Black house-holds, Hispanic households, house-holds whose reference person wasnot a high school graduate, andunmarried households with chil-dren.15
Growth in at least some types ofmaterial well-being measures was
6 U.S. Census Bureau
Table 2.Households With Income Below Poverty: Percentage With Selected Measures of MaterialWell-Being: 1992 and 1998
Indicator 1992 1998Percentage-
point change
Stove . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97.1 96.7 –0.4Enough of food wanted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.1 92.4 1.3Telephone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81.1 87.0 *5.9No smoke or odors in neighborhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89.7 93.2 *3.6No unpaid rent or mortgage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81.4 86.8 *5.4No unmet need for doctor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84.2 85.9 1.7No roof or ceiling leaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85.5 89.1 *3.6No abandoned houses in neighborhood. . . . . . . . . . . . . . . . . . . . . . . . . . . 83.3 88.3 *5.1No unmet essential expenses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68.4 69.3 0.9Air conditioner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.6 67.7 *15.1Dishwasher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.0 25.7 *5.7Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 18.4 *11.4
* Percentage-point change from 1992 to 1998 is statistically significant from 0 at the 90-percent confidence level.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
11 Among the factors used to describehouseholds is “household type” (see Table 3).This factor includes the traditional CensusBureau classifications of family or nonfamilyhouseholds (based on the presence of morethan one person related by birth or marriage)and classification based on the presence orabsence of children under 18 in the house-hold. Earlier reports found that the presenceof children was strongly related to lowermaterial well-being.
12 No generally accepted method exists tosummarize indicators of material well-beingin various domains. With the exception ofthe indicator of neighborhood safety, thesummary indicators used here were chosento be consistent with previous Census Bureaureports and to represent the full range of sub-jects covered by the individual measures(Bauman, 2003). Neighborhood safety waschosen because, among safety indicators, itwas close to the middle in terms of the1992–1998 change.
13 The difference in freedom from hous-ing problems between households withhouseholders 65 and older and householdswith householders aged 45 to 64 was notstatistically significant.
14 The difference between married house-holds without children and married house-holds with children in the possession of acomplete set of appliances was not statisti-cally significant.
15 Because Hispanics may be any race, datain this report for Hispanics overlap slightlywith data for the Black population. Based ondata from Wave 8 of the 1996 panel of theSIPP, 4 percent of Black households were alsoHispanic. Data for Asians and PacificIslanders and for the American Indian andAlaska Native population are not shown inthis report because of their small sample sizein the 1998 panel of the SIPP. None of thedifferences between Blacks and Hispanics inthe four types of material well-being shown inTable 3 was statistically significant.
observed for all population sub-groups. With one exception,households with a reference per-son aged 65 or over, all subgroupsexamined here had greater free-dom from housing problems in1998 than in 1992. Measures ofthe possession of basic appliancesand having fewer than two difficul-ties meeting basic needs improvedfor half or more of the groups as
well.16 Perceptions of neighbor-hood safety, which did not change
for the population as a whole, alsodid not change for most (20 of 22)of the population subgroups exam-ined here. Perception of neighbor-hood safety increased for house-holds with a female referenceperson and those with a Hispanicreference person.
U.S. Census Bureau 7
Table 3.Percent of Households With Favorable Measures of Material Well-Being by Characteristicsof Householder and Household Type: 1992 and 1998
Household characteristicPossess basic durables No housing problems
Considerneighborhood
safe
Fewer than twodifficulties meeting
basic needs
1992 1998 Change1 1992 1998 Change1 1992 1998 Change1 1992 1998 Change1
All households
Total . . . . . . . . . . . . . . . . . . . . . 44.6 50.5 5.8 73.0 79.2 6.2 91.0 91.4 0.3 88.3 90.4 2.1
Age of householder
15 to 29 . . . . . . . . . . . . . . . . . . 32.6 35.8 3.2 69.1 78.0 8.8 87.5 89.6 2.1 81.9 85.1 3.330 to 44 . . . . . . . . . . . . . . . . . . 47.7 53.5 5.8 69.1 77.2 8.1 90.5 91.1 0.6 83.8 87.1 3.345 to 64 . . . . . . . . . . . . . . . . . . 52.0 56.4 4.4 74.3 79.7 5.4 92.5 91.5 –1.0 90.5 91.4 0.965 or older. . . . . . . . . . . . . . . . 37.1 45.0 7.9 79.9 82.3 2.3 92.3 92.7 0.3 97.0 97.2 0.2
Sex of householder
Male . . . . . . . . . . . . . . . . . . . . . 50.8 55.1 4.2 75.0 80.8 5.8 93.2 93.3 0.1 90.0 92.5 2.5Female . . . . . . . . . . . . . . . . . . . 33.3 44.9 11.6 69.2 77.2 8.0 87.1 89.0 1.9 85.3 87.9 2.6
Race and Hispanic originof householder
White . . . . . . . . . . . . . . . . . . . . 48.2 53.9 5.7 75.3 80.3 5.0 92.7 92.9 0.2 89.6 91.6 2.0White non-Hispanic . . . . . . 50.3 56.8 6.5 76.7 81.5 4.8 93.9 93.8 –0.1 90.3 92.4 2.1
Black. . . . . . . . . . . . . . . . . . . . . 17.4 28.3 10.8 56.0 72.9 16.9 78.6 80.6 2.0 78.2 81.8 3.7Other . . . . . . . . . . . . . . . . . . . . 45.7 44.2 –1.5 69.6 73.9 4.4 89.8 90.5 0.7 90.0 90.6 0.5
Hispanic (any race) . . . . . . . . 20.3 24.6 4.3 56.6 68.7 12.1 78.3 83.8 5.5 80.1 84.0 3.8
Householder education
Not high school graduate . . . 20.3 23.1 2.8 65.5 72.2 6.7 86.5 86.3 –0.1 84.9 86.0 1.1High school graduate . . . . . . 40.9 46.1 5.2 72.6 79.8 7.3 91.0 91.3 0.2 86.4 89.1 2.7Some college . . . . . . . . . . . . . 51.1 55.3 4.2 74.4 79.4 5.1 91.4 91.3 –0.1 87.3 89.7 2.4Bachelor’s or more . . . . . . . . 67.3 70.1 2.8 79.2 83.1 3.9 94.9 95.1 0.2 95.2 96.1 0.9
Household type
Nonfamily alone . . . . . . . . . . . 29.2 34.7 5.6 75.6 79.9 4.3 89.6 90.3 0.8 91.5 91.2 –0.3Nonfamily with others . . . . . . 36.3 41.9 5.6 69.1 78.1 9.0 88.0 91.1 3.0 84.8 88.7 3.9Married, no children2 . . . . . . . 58.3 65.1 6.8 80.7 84.3 3.6 94.9 94.3 –0.6 94.8 96.2 1.4Married with children2 . . . . . . 56.1 64.2 8.1 70.6 78.9 8.3 93.0 92.9 –0.1 86.0 90.4 4.4Unmarried, no children2 . . . . 32.7 37.8 5.1 69.5 75.6 6.1 87.5 90.5 3.0 87.5 89.2 1.7Unmarried with children2 . . . 25.0 29.9 5.0 55.4 67.4 12.1 82.2 83.4 1.1 71.6 75.0 3.4
1 Percentage-point change from 1992 to 1998.2 Children include all household members under 18.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
16 The change in the rate of possession ofconsumer durables was statistically signifi-cant for all but the following: householdswith a reference person aged 15 to 29,Hispanics, households with a reference per-son with less than a high school diploma,nonfamily households living with others, andunmarried households without children. Thechange in having fewer than two difficultiesmeeting basic needs was statistically signifi-cant for 12 of the 22 groups.
8 U.S. Census Bureau
.4
e lba T
)y
altu
O(er
utid
ne
pxE
yb
sel
cih
eV
dn
as
ec
nail
pp
Ad
etc
ele
Sf
opi
hsr
en
wO
gnitr
op
eR
)s
UC(
stin
Ur
em
us
no
Cf
ot
ne
c re
P2
00
2d
na
29
91
:eli
ce
D
elicederutidnepx
E
llarevO
019
87
65
43
21
evaw
or ciM
..2 991
.. .. .. ......................8.67
3.490.49
7.097.88
9.381.87
4.377.46
8.0 64.9 3
..20 02
................... ...... ...2.39
7.794.89
9.790.69
8.598.49
5 .496.09
1.981 .77
rota re
gi rfeR
..2 991
...... .. .. ..................7.89
9.997.99
9.995.99
3.992.99
8.890. 89
7. 799.4 9
..2002
............................2.99
9.998.99
8.999.99
6.999.99
5.990.99
0.993.59
rezee rF
..29 91
.................... .. .. ....8.23
3.240.04
6.930.63
2.234.13
6 .824.82
5.825.0 2
..2002
................... .........8.03
8.736. 43
9 .6 36.23
0. 138.82
5 .922.7 2
1.8 26 .12
laso
psid
ega
br aG
..29 91
.......................... ..3.73
8.669 .65
7.7 44. 34
2.837.13
4 .0 35.3 2
4.0 22 .4 1
..2002
............................0.74
3.176.56
6.061.35
4.948.34
1.938.53
1. 822.32
r eh sa
W..
2991............................
0.570.59
0.393.09
9.487.77
8.170.96
2.365.06
5.44..
2002... .........................
0.083.59
0. 399.98
5.789 .48
7.878 .6 7
6. 170. 86
0.4 5rey r
D..
29 91............................
5.864.39
3.193.78
6.288.37
1.669.06
8.257.64
4.03..
2 002........... .. .. .. ...........
1.771.59
2.291.98
9.585.28
7.674.37
1.763.26
0.74V
Tr
o lo
C..
2991............................
7.494 .99
2. 990 .99
3.892. 79
4 .695 .59
6.295.19
5 .77..
2 002................. .. .. .......
2 .897 .99
5. 991 .9 9
5. 999. 89
7 .899 .89
1.8 91.7 9
4 .2 9ret
up
mo
C..
29916 .8
2.61.5
1.4............................
6.818.54
5.733.62
7.222.61
7.31..
2 002......................... .. .
3.951.19
7.682.08
4.474.66
4.558.84
8. 833. 03
0.12s t
nen
op
mo c
dn
uo
S..
2 991............................
3.759.48
7.770.07
1.860.26
7.453 .05
6.343.53
5 .62..
2 002............................
5 .275.98
4.887.38
8.188. 67
0.276 .17
5.366.75
9 .93R
CV
..2 991
. ...........................1.86
6.390. 29
2.884.48
4 .777.17
2 .2 66. 05
2. 831.2 2
..2002
............................9.68
6.796.69
6.596.39
8.291.98
9.988. 28
0. 579.55
d en i
bm
ocse v
o tS
1
..2991
............................0.89
8.997.99
7.993.99
1.990.99
7.897.79
7.795 .98
..2002
............................3.89
4.993.99
9.894.99
2.991.99
0 .993.89
0.991 .19
deni
bm
ocs re
hsaw
hsiD
2
..2 991
5. 11.............. ..............
7.848.78
6. 777.5 6
6. 953 .1 5
4.446 .7 3
2.9 20.2 2
..2 002
.............. ..............1.85
1.883 .38
9.5 72. 96
3 .2 68.45
3 .849.9 3
7.5 38.3 2
nav, kc
urt,
otu
A..
2991............................
1.580.69
9.699.59
7.594.39
4.197.7 8
1. 088. 86
6.54..
2002............................
7.586.49
8.592.59
3.495.39
4.195.98
7.181.37
0.84
1. sevo ts
r eh todna
,s evotsc irtc ele
,s evo tssa g
s edu lcnide nib
m ocsev ot
S2
.srehsaw hsid
e lbatropdna
srehs awhsid
ni-tliubsedulcni
den ibmoc
srehs awhs i
Do
noi tinif edsya lt uo
n agnis u
ybde ta erc
erase lice d
er uti dn epxE
.strahcevo ba
ehtn i
ded ulcn ie ra
st inula tne rni
ded ivorper a
tahtsecna ilp p
A:se toN
- fidyra
mir pe h
T.s erutidne pxe
f hca orppasyaltuo
ehT
.snoitinifede gagtro
me
mohdna
elciheveh t
niera
syaltu odna
, at addehs ilb up
EC
nides u
s a,ser utidne pxelatot
n eew te b
sec nere fecirp
sel aselci hev
secalpertidnepxelatot
EC
ehtelih
w,syaltuo
nidedulcni
erastne
myaplapicnir pegagtro
me
moH
. doir epecnerefer
yevruseht
gnirudeda
mstne
myapelcihev
htiw
esehtsredisnoc
noitinifedseru
wrofdeirrac
sidna
weivretnitsrif
s’U
Ceht
gniruddeirotnevni
sipihsren
woecnailppa
nonoita
mrofnieh
T.
mehtedulcni
tonseod
dnastne
mtsevnieb
ota
fI.s
weivretnitneuqesbus
otdra
rifeht
,ylralimi
S.detadpu
tonsi
noitamrofni
deirotnevnieht
,weivretni
tn euqesbusa
ni)n
woton
didylsu oiverp
tihcih
w(ecnailppa
nasesahcrup
UC
sinoita
mrofniecnailppa
weivretnits
.weivretn i
tne uqesb usa
nided racsi d
rodlos
siecnail ppa
naf i
de tadp uton
. 20 02,2 99 1
a taD
weiv retnIyevr u
Serutidn epx
Ere
mus noC
de hsil bup nu,scitsita t
Sr oba L
fouaer u
B:ecru o
S
Changes in indicators ofmaterial well-being ofpopulation subgroups
Although the rate of growth of indi-cators of material well-being variedsomewhat across subgroups, theperiod from 1992 to 1998 did notproduce great changes in thegroups’ relative standing. Of thefew population subgroups that hadlarger increases than others, mosthad had lower material well-beingthan the comparison groups did in1992. This pattern was true ofBlacks and Hispanics, two groupsthat made gains relative to non-Hispanic Whites in freedom fromhousing problems.17 Hispanics also
had a larger gain in perceivedneighborhood safety than did non-Hispanic Whites. Similarly, house-holds with a female reference per-son had a larger gain in theirpossession of basic appliances thanhouseholds with a male referenceperson, after starting with a lowerlevel of possession in 1992.18 Theshare of married households withchildren with fewer than two diffi-culties meeting basic needsincreased more than the shares ofnonfamily households living aloneand married households withoutchildren. Both married andunmarried households with children
experienced larger growth infreedom from housing problemsthan did single-person householdsand married households withoutchildren.
Smaller increases in material well-being indicators occurred in twogroups that already had higher lev-els. Gains in freedom from housingproblems and in having fewer thantwo difficulties meeting basic needswere smaller for households with areference person aged 65 or overthan for most younger groups.Households whose reference personhad a bachelor’s degree or moreeducation had smaller gains in free-dom from housing problems thanhouseholds whose reference personhad a high school diploma.
One notable exception was found inthe pattern of greater increases inmaterial well-being occurring ingroups with previous low measuresof material well-being. Householdswith working-age adults reportingthat they are limited in their workdue to disability have been found inthe past to have low levels of mate-rial well-being. This group contin-ued to have low well-being levelseven after controlling for income,assets, and other economic anddemographic factors (Bauman,2004). Similarly, the relative posi-tion of households with disabledmembers did not improve in any ofthe four areas examined here, andthey fell behind in possession ofbasic appliances and in consideringtheir neighborhoods safe (Figure 2).
In 1992, 36 percent of householdswith disabled working-age adults(aged 25 to 64) reported having afull set of appliances, comparedwith 46 percent of other house-holds. In 1998, the proportionswere 37 percent (an increase of 1percentage point) and 52 percent(an increase of 6 percentagepoints), respectively.
U.S. Census Bureau 9
Figure 2.Indicators of Material Well-Being by Household Disability Status: 1992 and 1998
Note: Households classified as disabled have one or more working age adults (aged 25 to 64) reporting that a health condition limits the work they can do.Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
Fewer than two difficulties meetingbasic needs - Not disabled
Fewer than two difficulties meetingbasic needs - Disabled
Consider neighborhood safe -Not disabled
Consider neighborhood safe -Disabled
No housing problems -Not disabled
No housing problems -Disabled
Possess basic appliances -Not disabled
Possess basic appliances -Disabled
19921998(In percent)
32.7
89.6
90.5
78.4
77.7
91.5
91.9
87.6
84.2
74.3
79.9
62.3
67.0
46.2
54.3
32.8
17 Blacks were not statistically differentfrom Hispanics in their rate of change infreedom from housing problems.
18 However, this finding may reflect agreater percentage of households identifyingwomen as the reference person in 1998 thanin 1992, possibly due to changes in inter-view procedures.
10 U.S. Census Bureau
.5
elba
T:
elic
eD
)y
altu
O(er
utid
ne
pxE
yb
)U
C(ti
nU
re
mu
sn
oC
re
pd
en
wO
sel
cih
eV
dn
as
ec
nail
pp
Ad
etc
ele
Sf
or
eb
mu
Ne
gar
ev
A2
00
2d
na
29
91
elicederutidnepx
E
llarevO
019
87
65
43
21
evaw
or ciM
..2 991
8.00.1
0.19.0
9.09.0
8.07.0
7.06 .0
4 .0.. .. .. ...................... ..
20 020.1
0.10.1
0.10.1
0.10.1
0.19 .0
9 .08 .0
......................... ...r
ota regi rfe
R..
2 9911.1
2.11.1
1.11.1
1.11.1
0 .10.1
0.10 .1
...... .. .. .................. ..2002
1.12.1
1.11.1
1.11.1
1.10.1
0.10.1
0 .1............................ rezee r
F..
29 913.0
5.04.0
4.04.0
3.03.0
3. 03 .0
3 .02 .0
.................... .. .. .... ..2002
3.04.0
4.04. 0
3 .03. 0
3.03 .0
3 .03 .0
2.0............................
laso
psid
ega
br aG
..29 91
4.07.0
6.05 .0
4.04. 0
3.03 .0
2 .02 .0
1 .0.......................... .. ..
20025.0
7.07.0
6.05.0
5.04.0
4.04.0
3.02.0
............................ r eh sa
W..
29 918.0
0.19.0
9.08.0
8.07.0
7. 06 .0
6 .04.0
............................ ..2002
8.00.1
9.09.0
9.09.0
8.08.0
7.07.0
5.0............................
reyrD
..29 91
7.09.0
9.09.0
8.07.0
7.06.0
5.05.0
3.0............................ ..
2 0028.0
0.19.0
9.09.0
8.08.0
7.07.0
6.05.0
............ .. .. ............ VT
ro l
oC
..2991
6.14.2
1.29. 1
8 .16.1
5.14 .1
3 .12 .1
9 .0............................ ..
2 0021 .2
9 .26.2
5. 24 .2
2. 20 .2
9 .18 .1
6 .13 .1
.................. .. ........ret
up
mo
C..
29912.0
5.04.0
3.02.0
2.01.0
1 .01.0
1.00.0
............................ ..20 02
8.04.1
2.10.1
9.08.0
6.06 .0
4.03.0
3.0.......................... ..
s tne
no
pm
ocd
nu
oS
..2991
7.02.1
0.19.0
8.07.0
6.06. 0
5 .04.0
3.0............................ ..
20 020.1
5.13 .1
2.11.1
0.19.0
9 .08 .0
7 .05 .0
............................R
CV
..2 991
8.04.1
2.11.1
0.19.0
8 .07 .0
5.04 .0
2 .0. ........................... ..
20024.1
0.28.1
7.15.1
4.13.1
2 .11.1
9.06.0
............................d e
n ib
moc
se vo t
S1
..2991
1.12.1
1.11.1
1.11.1
1.11. 1
0.10.1
9.0............................ ..
20020.1
2.11.1
1.11.1
0.11.1
0. 10 .1
0 .19.0
............................de
nib
moc
s rehsa
whsi
D2
..2 991
5 .09 .0
8.07. 0
6.05. 0
4 .04 .0
3 .02 .0
1 .0.............. .............. ..
2 0026.0
9.08.0
8 .07.0
6. 06 .0
5 .04 .0
4 .02 .0
................ ............nav
, kcurt
,ot
uA
..2991
6.15.2
3.21.2
9.17.1
6.14 .1
2.19.0
6 .0............................ ..
20027.1
4.23.2
1.20.2
8.16.1
5.12.1
0.16.0
............................1
.sevotsreh to
dn a,sevots
cirtcel e,sevo ts
sagsed ulcn i
d eni bmoc
sevotS
2.sr ehsa
whsidelba trop
dnasrehsa
whsidni-tliub
sedulcnidenib
mocsr ehsa
whsiD
onoitinifed
syaltuona
gnisuyb
detaercera
selicederutidnepx
E.str ahc
evobaeht
nidedulcni
erastinula tner
n idediv orp
erataht
secnailppA:se to
N-fid
yramirp
ehT
.serutidnepxef hcaorppa
syaltuoeh
T.snoitinifed
egagtrom
emoh
dnaelcihev
ehtn i
e rasy al tuo
dn a,atad
d ehsi lbupE
Cn i
desusa
,s eru tid nepxel atotnee
wtebsecneref
ecirpselas
elcihevsecalper
tid nepxel atotE
Ceht
el ihw
,sy al tu oni
de du lc nier a
st nem ya pl apic nirp
eg agtrom
em o
H.d oire p
ecner efe ry evrus
eh tgn irud
ed am
s tn emyap
e lcih evh ti
wes eht
sr edisnocnoit ini fed
ser uwrof
dei rracsi
dnaweivretn i
tsrifs’
UC
eh tgnirud
deirotnev nisi
pi hsrenwo
ecna ilp pano
no it amrofni
e hT
.m eht
edulc nit on
s eodd na
s tn emts evni
ebo t
afI
.sweivretni
tneuqesbusot
drarif
eht,ylrali
miS
.detadputon
sinoita
mrofnideirotnevni
eht,
weivretnitneuqesbus
ani
)nwo
tondid
ylsuoiverpti
hcihw (
ecnailppana
se sahcrupU
Csi
noitamrofni
ecnailppaweivretni
ts.
w eivretn itne uqesb us
ani
de dracsi dro
dlossi
ecn ail ppana
f ide tadp u
ton. 20 02
,2 99 1ata
Dwe ivretnI
yevr uS
erutid nepxE
remus no
Cde hsi lbup nu
,scitsita tS
roba Lfo
uaer uB
:ecru oS
Similarly, households with disabledworking-age adults fell behind intheir perception of neighborhoodsafety by 3 percentage points,while other households gained by1 percentage point. Larger propor-tions of households with disabledadults reported freedom fromhousing problems and fewer thantwo difficulties meeting basicneeds. Those improvements wereabout the same as the improve-ments for other households.
TRENDS IN THE POSSESSIONOF CONSUMER DURABLES:1992 TO 2002
Data from the U.S. ConsumerExpenditure (CE) Survey also shedlight on material well-being.Consumer units (CUs), rather thanhouseholds or families, are theunit of analysis.19 Tables 4 and 5and Figures 3, 4, and 5 includeinformation on the percentage ofconsumer units reporting the own-ership of selected appliances andvehicles and the average numberof these durables in 1992 and2002. Tables 6 and 7 show meas-ures of their distribution over thefull 11-year period, 1992 to 2002.The information presented is basedon appliance ownership collectedduring the first interview.20
U.S. Census Bureau 11
.6
e lba T
20
02
ot2
99
1:
sel
cih
eV
dn
as
ec
nail
pp
Ad
etc
ele
Sf
os
eta
Rpi
hsr
en
wO
fo
sex
ed
nIn
oitart
ne
cn
oC
)sy altuoeru ti dnepx e
yB(
sexedninoitartnecno
C
20021002
00029991
89917991
69915991
4991399 1
299 1
......rotare gir fe
R..................
10.000.0
00.010.0
10.010.0
00.010.0
10.010. 0
10.0. ....
VT
rol oC
............... .......10.0
10.010.0
20.020.0
20.020.0
20. 020.0
2 0.03 0. 0
.eva
worc iM
....... .................30.0
40.040.0
50.060.0
60.060.0
80.090.0
01 .021 .0
............... .. .............R
CV
70.070.0
80.090.0
01.011.0
11.031.0
51.071 .0
91.0r ehsa
W.. .. .. ......................
80.090.0
11.001.0
01.011.0
11.011 .0
11.011.0
21 .0......reyr
D........ .. .. ............
01.011.0
31.031.0
31.031.0
41.041. 0
51 .061.0
61.0st neno p
mocdn uo
S................ .
11. 02 1.0
21. 021 .0
41. 041.0
5 1. 061. 0
71 .081.0
81.0denib
m ocsev ot
S1
.. .. .. ............10 .0
10.020.0
10.010.0
10. 010.0
10 .010.0
10.010.0
.. ....retupm o
C....................
22 .032 .0
52.072.0
0 3.023. 0
43 .083 .0
73.083.0
04 .0de nib
m ocsr ehsa
whsiD
2.............
02. 01 2. 0
12.01 2. 0
32 .03 2.0
3 2. 052.0
62 .062 .0
72 .0.lasops id
egabraG
. .. .. .............81 .0
02.081 .0
8 1. 012.0
1 2.091.0
32.012 .0
22 .042 .0
rezee rF
1 1. 01 1.0
... ...... ...... ...... ...... .8 0. 0
1 1. 011.0
21.011 .0
11.01 1. 0
11.021.0
..nav
,kcurt,otu
A. .. .. .. .. .. .......
80.070.0
70 .080.0
80. 080.0
80.080.0
80.080.0
90.0
1.s evo ts
r eht odna
,sev otsci rtcele
,s evotssa g
sed ul cnid enib
m ocs evot
S2
.s rehsawhsi d
elbatro pdna
srehs awhs id
ni-tl iubse dulcni
den ibmoc
srehsawhsi
D
tila uq enion
sier eh t
nehw
or ezfo
eu lava
eva hll iw
xe dnino ita rt necnoc
A.tn eiciff eocin iG
eh tfo
g ola naetair avib
aer a
sex edn inoit artnec no
C:seto
N-ucitr ap
afo
pihs renw o
ehtn i
yrcalauqe
eraelbarud
ralucitrapa
gninwo
stinure
musnocf o
se gatnecr epeht
,sdrow
r ehtonI
. noi tubirtsidsya lt uo
erutidnepx eeht
ss orcae lbaru d
r alrehgih
fosexednI
.selicedeht
ssoihsren
wofo
noitartnecnoca
etacidniorez
nahtssel
sexedniel ih
w,syaltuo
rehgihfo
st inure
musnocrovaf
syaltuoni
ytilau qe nita ht
et aci dniseul av
rewol
htiw
stinure
musnocgno
map
.sya lt uoeruti dnepx elat ot
ybdeknar
eras tin u
remusnoc
,sexed nies eht
roF
. syaltu o
. 2 002, 2991
at aD
we ivre tn Iy evru
Seru ti dnepx
Ere
m usnoC
dehsi lbu pnu,sc its it at
Srob aL
foua eru
B: ec ruo
S
19 A consumer unit consists of members ofa household related by birth, marriage, adop-tion, or some other legal arrangement; a sin-gle person living alone or sharing a house-hold with others, but who is financiallyindependent; or two or more people livingtogether who share responsibility for at leasttwo out of the three major types of expenses:food, housing, and other expenses. Also, stu-dents living in university-sponsored housingare included in the sample as separateconsumer units.
20 The information on appliance ownershipis inventoried during the consumer unit’s (CU)first interview and is carried forward to subse-quent interviews. If a CU purchases an appli-ance (which it previously did not own) in asubsequent interview, the inventoried infor-mation is not updated. Similarly, the firstinterview appliance information is notupdated if an appliance has been sold or dis-carded by the time of a subsequent interview.An appliance that is provided in a rental unitis included and identified as “owned” althoughthe consumer unit does not own the item buthas exclusive access to it.
As presented in the 2003 report onsupplemental measures, results areshown by deciles. First, consumerunits are ranked by expenditureoutlays from lowest to highestexpenditure level and then dividedinto 10 equal groups of CUsaccording to this ranking.Percentages of ownership and theaverage number of durables ineach decile are then produced.Expenditure deciles are created byusing an outlays definition, as inthe 2003 report. The primary dif-ferences between total expendi-tures used in CE published dataand outlays used in the tables andcharts in this report are in the vehi-cle and home mortgage defini-tions. The outlays approachreplaces vehicle sales price withvehicle payments made during thesurvey reference period. Homemortgage principal payments areincluded in outlays, while the CEtotal expenditures definition con-siders them to be investments anddoes not include them.21
As shown in the 2003 report onsupplemental measures of well-being (U.S. Census Bureau, 2003)and by others (noted previously aswell), the majority of consumerunits own or have access to most ofthe appliances and vehicles invento-ried in the CE (exceptions are freez-ers and garbage disposals, and in
1992, computers and dishwashers).The percentages of ownership forrefrigerators, stoves, and color tele-visions are consistently high overthe deciles from 1992 to 2002. Thegreatest gain in ownership from1992 to 2002 is for computers. In1992, 18.6 percent of all consumerunits owned or had access to acomputer in their homes; by 2002,
59.3 percent did (Table 4 and Figure 3). Consumer units in thehigher deciles were more likely thanthose in lower deciles to own com-puters (Table 4 and Figure 5). Thistrend is also exhibited by the con-centration indexes shown in Table6: the higher the concentrationindex, the greater the concentrationof computers in the higher
12 U.S. Census Bureau
Figure 3.Ownership of Selected Appliances and Vehicles by Consumer Units: 1992 and 2002
1 Stoves combined includes gas stoves, electric stoves and other stoves.Dishwashers combined includes built-in dishwashers and portable dishwashers. Source: Bureau of Labor Statistics, Consumer Expenditure Survey interview data, 1992 and 2002.
Freezer
Garbage disposal
Dishwashers combined1
Computer
Sound components
Dryer
Washer
VCR
Microwave
Auto, truck, van
Color TV
Stoves combined1
Refrigerator
19922002
99.2
32.8
30.8
37.3
47.0
48.7
58.1
18.6
59.3
57.3
72.5
68.5
77.1
75.0
80.0
68.1
86.9
76.8
93.2
85.1
85.7
94.7
98.2
98.0
98.3
98.7
(In percent)
21 The updated values for vehicle owner-ship for 2001 are different than thosereported in the 2003 report on supplementalmeasures of well-being (U.S. Census Bureau,2003). The values reported here were pro-duced by first ranking all consumer units bytotal expenditure outlays. Then the owner-ship rates and average number of durablesowned in each decile of consumer units wereproduced. The 2001 results for vehiclespublished in the 2003 publication werebased on ranking consumer units by expen-ditures using a family size adjustment. Theappliance results for 2001 are slightlydifferent from those reported in the 2003report because of a slight modification tothe construction of the ranking and thecreation of deciles.
expenditure outlay deciles.22 Theconcentration index has been
decreasing over time, indicating anincreasingly more equal
distribution. The average numberof computers owned also increasedfrom 1992 through 2002 (Table 5).In the top deciles, each consumerunit had on average more than onecomputer in 2002; in 1992 it wasless than one.
Other large increases in ownershiprates occurred for microwaveovens and entertainment equip-ment such as VCRs and soundcomponents. Microwaves werealmost as prevalent as refrigeratorsby 2002 (Table 4), and their distri-bution became fairly equal over thedeciles as evidenced by the con-centration indexes decreasing from0.12 to 0.03 from 1992 to 2002(Table 6). The average number ofmicrowaves in a consumer unit wasone or almost one for each decileby 2002.
Entertainment equipment, repre-sented by VCRs and sound compo-nents, increased in ownership andprevalence over the time period.About 87 percent of all consumerunits in 2002 owned a VCR, com-pared with 68 percent in 1992(Table 4). The 1992 and 2002 per-centages for sound componentswere slightly lower than the per-centages for the same years forVCRs. The average number ofvideocassette recorders owned inthe upper deciles doubled to abouttwo per consumer unit in 2002,from less than two in 1992 (Table5). The number of sound compo-nents was 1.5 in the upper decilein 2002. The concentration indexesreveal that the ownership of VCRsvaried along the expenditure out-lays distribution but that consumerunits all along the distributionwere becoming more likely to owna VCR in 2002 (index = 0.07) com-pared with 1992 (index = 0.19)(Table 6). Sound components weresomewhat concentrated amongconsumer units in the upperdeciles (Table 4) but were
U.S. Census Bureau 13
Figure 4.Average Number of Selected Appliances and Vehicles Owned by Consumer Units: 1992 and 2002
1 Stoves combined includes gas stoves, electric stoves and other stoves.Dishwashers combined includes built-in dishwashers and portable dishwashers. Source: Bureau of Labor Statistics, Consumer Expenditure Survey interview data,1992 and 2002.
Freezer
Garbage disposal
Dishwashers combined1
Computer
Sound components
Dryer
Washer
VCR
Microwave
Auto, truck,van
Color TV
Stoves combined1
Refrigerator
19922002
1.1
0.3
0.3
0.4
0.5
0.5
0.6
0.2
0.8
0.7
1.0
0.7
0.8
0.8
0.8
0.8
1.4
0.8
1.0
1.6
1.7
1.6
2.1
1.1
1.0
1.1
22 A concentration index is a bivariateanalog of the Gini coefficient. The index isbased on the distribution of one variable incombination with another variable. Forexample, in Table 6, the concentration indexrepresents a particular relationship betweenthe distribution of the ownership of adurable and the rank of consumer units bytotal expenditure outlays. Graphically theindex can be represented by the plot of thecumulative distribution of consumer unitsowning a particular durable on the y axis byconsumer units ranked by total expenditure
outlays on the x axis. A concentration index will have a value of
zero when there is no inequality in the own-ership of a particular durable across theexpenditure outlays distribution. In otherwords, the percentages of consumer unitsowning a particular durable are equal acrossthe deciles. Indexes of higher values indicatethat inequality in outlays favor consumerunits of higher outlays, while indexes lessthan zero indicate a concentration of owner-ship among consumer units with loweroutlays.
becoming more equally distributed(Table 6) (2002 index = 0.11 ver-sus 1992 index = 0.18).
On average, by the year 2002,most consumer units owned or hadaccess to only one each of thedurables with a few exceptions(Table 5). The average was aboutone-half for food freezers, garbagedisposals, and dishwashers. Theaverage overall ranged from 1.4 to2.1 for VCRs (already discussed),color televisions, and vehicles. Theaverage number of color televi-sions and vehicles was greaterthan one for most deciles, with theaverage for color televisionsgreater over the deciles and overtime. The average number of vehi-cles owned changed very little overthe 1992 to 2002 period. The dis-tribution of vehicles was more con-centrated among consumer unitswith greater expenditures thanamong those with lower expendi-tures. Compared with vehicles, thepossession of color televisions wasrelatively more equally distributedacross the total expenditure distri-bution (Table 7).
ENERGY USE AND POVERTY:1981 TO 2001
The Energy Information Administra-tion (EIA) of the U.S. Department ofEnergy conducts the ResidentialEnergy Consumption Survey (RECS)every four years. The RECS collectsdata on energy consumption andexpenditures and energy-relatedsubjects for the household sector ofthe U.S. economy.
Between 1981 and 2001, the totalnumber of households representedby the RECS increased by 28.8 per-cent, from 83.1 million in 1981 to107.0 million in 2001 (Table 8).Over the same period, the numberof households with householdincome below the poverty lineincreased by 36.4 percent, from
14 U.S. Census Bureau
Figure 5.Ownership of Selected Appliances and Vehiclesby Consumer Units at the First and TenthExpenditure (Outlay) Decile: 1992 and 2002
¹ Stoves combined includes gas stoves, electric stoves and other stoves.Dishwashers combined includes built-in dishwashers and portable dishwashers. Note: Expenditure deciles are created by using an ‘‘outlays’’ definition of expenditures. Source: Bureau of Labor Statistics, unpublished Consumer Expenditure Survey Interview Data 1992, 2002.
First (lowest) expenditure decileTenth (highest) expenditure decile
(In percent)
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
1992
2002
Refrigerator
Stoves combined¹
Color TV
Auto, truck, van
Microwave
VCR
Washer
Dryer
Sound components
Computer
Dishwashers combined¹
Garbage disposal
Freezer
94.999.9
95.399.9
89.599.8
91.199.4
77.599.4
92.499.7
45.696.0
48.094.6
39.494.3
77.197.7
22.193.6
55.997.6
44.595.0
54.095.3
30.493.4
47.095.1
26.584.9
39.989.5
4.145.8
21.091.1
11.587.8
23.888.1
14.266.8
23.271.3
20.542.3
21.637.8
11.0 million in 1981 to 15.0 mil-lion in 2001.
Total annual Btu consumption perhousehold decreased by 17.5 per-cent between 1981 and 2001 froman average of 111.7 million Btu to92.2 million Btu. Among thosehouseholds with householdincomes below the poverty line,total Btu consumption was lowerand the decrease between 1981and 2001 was larger. Total Btuconsumption per household forthese households decreased by23.7 percent between 1981 and2001 from an average of 92.7 mil-lion Btu to 70.7 million Btu.Among households with incomesabove 1.25 times the poverty level,total annual Btu consumption perhousehold decreased by 16.9 per-cent between 1981 and 2001, froman average of 115.8 million Btu to96.2 million Btu.
After adjusting for inflation, totalannual energy expenditures amongall U.S. households decreased by16.9 percent between 1981 and2001, from an average of $1,760to $1,463 in 2000 dollars (Figure6). Among the households withincomes below the poverty line,total energy expendituresdecreased by 22.1 percent, froman average of $1,431 to $1,115.Among households with incomesabove 1.25 times the poverty level,total energy expendituresdecreased by 16.5 percent, from$1,832 to $1,529.
Home heating
Paralleling total annual space heat-ing Btu consumption, total annualinflation-adjusted space heatingexpenditures decreased by 33.5percent between 1981 and 2001,from an average of $705 to $469(Table 9). Among households inpoverty, total inflation-adjustedspace heating Btu expenditureswere lower and the decrease
U.S. Census Bureau 15
.7
e lba T
20
02
ot2
99
1:
de
nw
Os
elci
he
Vd
na
se
cn
ailp
pA
det
cel
eS
fo
re
bm
uN
eg
are
vA
fo
sex
ed
nIn
oitart
ne
cn
oC
)sy altuoeru ti dnepx e
yB(
sexedninoitartnecno
C
20021002
00029991
89917991
69915991
4991399 1
299 1
......rotare gir fe
R..................
30.030.0
30.030.0
30.030.0
30.030.0
30.030. 0
30.0. ....
VT
rol oC
............... .......31.0
31.041.0
41.041.0
31.041.0
51. 051.0
5 1.05 1. 0
.eva
worc iM
....... .................40.0
50.050.0
60.060.0
70.070.0
90.001.0
01 .021 .0
............... .. .............R
CV
71.071.0
71.071.0
81.091.0
91.002.0
12.022 .0
42.0r ehsa
W.. .. .. ......................
90.001.0
11.001.0
01.011.0
11.011 .0
11.011.0
21 .0......reyr
D........ .. .. ............
01.011.0
31.031.0
31.031.0
41.041. 0
51 .061.0
61.0st neno p
mocdn uo
S................ .
7 1. 09 1.0
81. 091 .0
02. 002.0
0 2. 022. 0
12 .022.0
22.0denib
m ocsev ot
S1
.. .. .. ............30 .0
30.030.0
30.030.0
30. 020.0
30 .030.0
30.030.0
.. ....retupm o
C....................
72 .082 .0
03.023.0
4 3.053. 0
73 .014 .0
93.004.0
24 .0de nib
m ocsr ehsa
whsiD
2.............
02. 01 2. 0
12.00 2. 0
32 .03 2.0
3 2. 062.0
62 .062 .0
72 .0.lasops id
egabraG
. .. .. .............91 .0
02.091 .0
8 1. 012.0
1 2.002.0
32.022 .0
32 .042 .0
rezee rF
1 1. 01 1.0
... ...... ...... ...... ...... .8 0. 0
1 1. 011.0
21.001.0
11.01 1. 0
21.021.0
.selcih e
V..........................
91.081.0
81.091.0
91. 091.0
02.091.0
02.002.0
12.0
1.s evo ts
r eht odna
,sev otsci rtcele
,s evotssa g
sed ul cnid enib
m ocs evot
S2
.s rehsawhsi d
elbatro pdna
srehs awhs id
ni-tl iubse dulcni
den ibmoc
srehsawhsi
D
tilauqenion
siereht
nehw
orezfo
eulava
evahlliw
xedninoitartnec noc
A. tne ici ffeo ciniG
eh tfo
go lanaetairavib
aera
sexednino itart nec no
C:seto
N-ucitrap
afo
pihsrenwo
ehtni
yrca lauqe
e rae lbar ud
ralucitrapa
gninwo
s tinure
musn ocfo
seg atn ecrepeht
,s drow
rehtonI
.noi tubi rts idsyaltuo
e rut idn epxee ht
ss orcaelb arud
ra lrehg ih
fosex ednI
.s elicede ht
ss oihsr en
w ofo
noitar tn ecn oca
et acid niorez
na htss el
sex edn ie lih
w,sya ltu o
r ehg ihfo
stin ure
musnocrov af
sya ltu oni
yt ila uqe nit aht
etacidniseulav
rewol
ht iw
stin ure
musno cgno
map
.syaltuoer ut idn epxela tot
y bd eknar
er as ti nu
remus noc
,sex ednies eht
roF
.sya lt uo
.2002,2991
ataD
weivretnIyevru
Serutidnepx
Ere
musnoC
dehs ilbupnu,scitsitat
SrobaL
f ouaeru
B:e cruo
S
between 1981 and 2001 waslarger. The average total annualspace heating expenditures forthese households decreased by37.7 percent from $608 to $379.Among households with incomesabove 1.25 times the poverty level,total space heating expendituresper household decreased by 32.9percent between 1981 and 2001from an average of $723 to $485.
Two space heating energy con-sumption and expenditures meas-ures providing insight on the“energy burden” for a housing unitare Heating Btu Intensity andHeating Dollar Intensity. HeatingBtu Intensity is a measure of thenumber of Btu consumed per1,000 square feet of heated livingspace and heating degree-day.23
Heating Dollar Intensity is a meas-
ure of the dollars spent per 1,000square feet of heated living spaceand heating degree-day.
Heating Btu Intensity for all U.S.housing units has been decreasingsince 1981, when it was 9.0, to its2001 low of 6.4. These decreaseswere observed for all incomegroups (Table 9). Heating BtuIntensity was highest for house-holds whose income was below the
poverty level, ranging between 8.7and 12.8,24 and lowest for thosehousing units whose householdincome was greater than 1.25times the poverty level, rangingbetween 6.1 and 8.6. Clearly, the“energy burden” is greater forhouseholds below the poverty linethan for more affluent ones.
Heating Dollar Intensity for all U.S.housing units decreased from itshigh in 1981, when it was 9.6, to itslow of 6.3 in 1993.25 In 2001,Heating Dollar Intensity increased to6.9. This pattern of decreases
16 U.S. Census Bureau
Table 8.Energy Consumption and Expenditures by Selected Household Characteristics: 1981 and2001
Characteristic
1981 2001
Number ofhouseholds
(millions)
Total Btus1
per household(thousand Btus)
Total dollarsper household
(inflation-adjusted
U.S. dollars)
Number ofhouseholds
(millions)
Total Btus1
per household(thousand Btus)
Total dollarsper household
(inflation-adjusted
U.S. Dollars)
All households . . . . . . . . . . . . . . . . . . . . . 83.1 111,720 1,760 107.0 92,202 1,463
Household incomeLess than $10,000. . . . . . . . . . . . . . . . . 23.3 94,160 1,453 11.0 65,248 1,019$10,000 to $14,999. . . . . . . . . . . . . . . . 12.5 104,135 1,630 7.7 69,704 1,101$15,000 to $19,999. . . . . . . . . . . . . . . . 10.7 103,669 1,633 8.9 80,479 1,265$20,000 to $29,999. . . . . . . . . . . . . . . . 17.7 114,696 1,848 14.0 83,380 1,289$30,000 to $39,999. . . . . . . . . . . . . . . . 10.1 132,567 1,986 13.9 86,881 1,370$40,000 to $49,999. . . . . . . . . . . . . . . . 4.3 130,480 2,152 13.2 92,804 1,488$50,000 to $74,999. . . . . . . . . . . . . . . . 3.1 149,313 2,521 21.7 102,516 1,650$75,000 or more . . . . . . . . . . . . . . . . . . 1.4 204,742 3,181 16.6 124,675 1,993
Income relative to poverty thresholdIncome below poverty threshold. . . . . 11.0 92,733 1,431 15.0 70,697 1,115Income 1 to 1.25 times poverty . . . . . 4.8 98,311 1,508 5.1 87,421 1,370Income above 1.25 times poverty . . . 67.3 115,779 1,832 86.9 96,192 1,529
1British thermal unit (Btu): The quantity of heat required to raise the temperature of 1 pound of liquid water by 1 degree Fahrenheit at thetemperature at which water has its greatest density (approximately 39 degrees Fahrenheit).
Notes: Household income is income from all sources, self-reported by the household respondent. Consumption and expenditures are formajor fuels used by the household, including electricity, natural gas, fuel oil, kerosene, and liquefied petroleum gas (LPG) as applicable.
Source: Energy Information Administration, Office of Energy Markets and End Use, Forms EIA-457 A-G of the 1981 Residential EnergyConsumption Survey and 2001 Residential Energy Consumption Survey.
23 A heating degree day is a measure ofhow cold a location is over a period of timerelative to a base temperature, most com-monly specified as 65 degrees Fahrenheit.The measure is computed for each day bysubtracting the average of the day’s highand low temperatures from the base temper-ature (65 degrees), with negative values setequal to zero. Heating degree-days aresummed to create a heating degree-daymeasure for a specified reference period.Heating degree-days are used in energyanalysis as an indication of space heatingenergy requirements or use.
24 The level of 12.8 for Heating BtuIntensity for households with income belowthe poverty threshold was reached in 1987.See “Additional Data” at the end of the textof this publication for information on how toaccess these data.
25 Information on 1993 is taken from tab-ulations that will be published separately bythe Energy Information Administration.
followed by an increase wasobserved for all income groups.Heating Dollar Intensity (afteradjusting for inflation) was highestfor households whose householdincome was below the poverty level,ranging between 9.7 and 12.3, andwas lowest for households whoseincome was greater than 1.25 timesthe poverty level, ranging between6.5 and 9.2.
Air-conditioning
In 1981, 47.6 million households,or 57.3 percent, used electricity forair-conditioning their homes. In2001 this number increased to80.8 million or 75.5 percent.26
Despite the fact that air-condition-ing Btu consumption per house-hold rose by 8.5 percent over the
1981–2001 period, annual perhousehold expenditures (afteradjusting for inflation) fell by 16.5 percent from $231 per year in1981 to $193 per year in 2001(see Table 10). This result reflectsthe fact that the cost, in inflation-adjusted dollars, of electricity, theprincipal source of energy for air-conditioning, decreased 22.5 per-cent over this period.
U.S. Census Bureau 17
Table 9.Home Space Heating Consumption and Expenditures by Selected HouseholdCharacteristics: 1981 and 2001
Characteristic
1981 2001
Dollars forspace heatingper household
(inflation-adjusted
U.S. dollars)
Heating Btu1
intensity(Btus perthousand
square feetand
degree-day2)
Heating dollarintensity
(inflation-adjusted U.S.
cents persquare feet and
degree-day2)
Dollars forspace heatingper household
(inflation-adjusted
U.S. dollars)
Heating Btu1
intensity(Btus perthousand
square feetand
degree-day2)
Heating dollarintensity
(inflation-adjusted U.S.
cents persquare feet and
degree-day2)
All households (that use eitherelectricity, natural gas, fuel oil,kerosene, or LPG for heating . . . . . . 705 9.0 9.6 469 6.4 6.9
Household incomeLess than $10,000. . . . . . . . . . . . . . . . . 656 11.4 12.4 372 8.7 9.8$10,000 to $14,999. . . . . . . . . . . . . . . . 670 9.3 10.2 389 7.9 8.8$15,000 to $19,999. . . . . . . . . . . . . . . . 692 8.7 9.6 439 7.9 8.5$20,000 to $29,999. . . . . . . . . . . . . . . . 695 8.2 8.8 447 7.9 8.5$30,000 to $39,999. . . . . . . . . . . . . . . . 745 8.2 8.0 448 7.2 7.7$40,000 to $49,999. . . . . . . . . . . . . . . . 758 7.4 8.1 479 6.2 6.7$50,000 to $74,999. . . . . . . . . . . . . . . . 887 8.1 9.0 488 5.6 5.9$75,000 or more . . . . . . . . . . . . . . . . . . 1,168 8.9 8.9 585 5.3 5.5
Income relative to poverty thresholdIncome below poverty threshold. . . . . 608 11.5 12.3 379 8.7 9.7Income 1 to 1.25 times poverty . . . . . 676 11.5 12.4 443 8.2 8.9Income above 1.25 times poverty . . . 723 8.6 9.2 485 6.1 6.5
1British thermal unit (Btu): The quantity of heat required to raise the temperature of 1 pound of liquid water by 1 degree Fahrenheit atthe temperature at which water has its greatest density (approximately 39 degrees Fahrenheit).
2Heating Degree-Days (HDD): A measure of how cold a location is over a period of time relative to a base temperature, most commonlyspecified as 65 degrees Fahrenheit. The measure is computed for each day by subtracting the average of the day’s high and low tempera-tures from the base temperature (65 degrees), with negative values set equal to zero. Each day’s heating degree-days are summed to cre-ate a heating degree-day measure for a specified reference period. Heating degree-days are used in energy analysis as an indication ofspace heating energy requirements or use.
Notes: Household income is income from all sources, self-reported by the household respondent. Consumption and expenditures are formajor fuels used by the household, including electricity, natural gas, fuel oil, kerosene, and liquefied petroleum gas (LPG) as applicable.
Source: Energy Information Administration, Office of Energy Markets and End Use, Forms EIA-457 A-G of the 1981 Residential EnergyConsumption Survey and 2001 Residential Energy Consumption Survey.
26 Unpublished tabulations from the RECS.
Annual per household expendi-tures for air-conditioning were pos-itively related to income relative tothe poverty level. Annual expendi-tures by households with incomesbelow the poverty level were lowerthan those by households withincomes greater than 1.25 timesthe poverty level. The 1981–2001decrease in expenditures amonghouseholds below the poverty levelwas 26.4 percent, compared with15.5 percent among more affluenthouseholds.
Cooling Btu Intensity for all U.S.housing units has been decreasingsince 1981, when it was 4.5, to its2001 low of 2.8. Decreases wereobserved for all income groups.Cooling Btu Intensity was higherfor households whose householdincome was below the povertylevel, ranging between 4.8 and 3.4,than for households whose house-hold income was greater than 1.25times the poverty level, rangingbetween 4.5 and 2.8.
As was the case with Cooling BtuIntensity, Cooling Dollar Intensity(after adjusting for inflation) alsodecreased from its high of 14.6 in1981 to its 2001 low of 7.1.These decreases were observed forall income groups. Cooling DollarIntensity (after adjusting for infla-tion) was highest for householdswith incomes below the povertylevel, ranging between 15.7 and8.2, and lowest for householdswhose household incomes weregreater than 1.25 times thepoverty level, ranging between14.7 and 7.0.
TWO DECADES OF CHANGE
This report has presented meas-ures of material well-being fromthree separate survey programs,covering the period from 1981 to2002. During the 1990s, improve-ments were observed in the pos-session of most types of consumer
durables, in most housing andneighborhood conditions, and inmeeting basic needs. Possessionof consumer durables continued toclimb through 2002.
Notable growth was observed inthe possession of newer goodssuch as microwaves, computers,and VCRs during the entire 1992to 2002 period. On the otherhand, no clear trend appeared inperceptions of crime and safetyduring the period 1992 to 1998.Some crime and safety indicatorsrose during the period while atleast one fell; people were morelikely to have stayed at home outof concern for safety in 1998 thanin 1992.
Nearly all groups participated inthe gains in possession of
consumer durables from 1992 to2002. During this period, the CEdata showed that the lowestexpenditure decile maintained orincreased its rate of possession of13 types of durables. According toSIPP data, groups defined by age,sex, race, Hispanic origin, educa-tion, and household type (marriedor unmarried, with or without chil-dren) increased their possession ofconsumer durables, reported fewerhousing problems, and had lessdifficulty meeting basic needsbetween 1992 and 1998. In con-trast, households with a disabledworking-age adult showed lowergains than the rest of the popula-tion in the possession of consumerdurables and freedom from thefear of crime.
18 U.S. Census Bureau
Figure 6.Total Inflation-Adjusted Energy Expenditures per Household by Income Relative to Poverty Threshold
Sources: Energy Information Administration: 1981, 1987, 1993, and 2001 Residential Energy Consumption Surveys.
1981198719932001
$1,529
$1,760
$1,522
$1,460
$1,463
$1,431
$1,274
$1,202
$1,115
$1,508
$1,330
$1,300
$1,370
$1,832
$1,579
$1,519
Income above1.25 times
poverty
Income1 to 1.25 times
poverty
Income belowpoverty threshold
All households
(Inflation-adjusted year 2000 dollars)
Lower income or expenditurehouseholds increased their posses-sion of many consumer durableitems. Quality of consumerdurables was not observed inthese surveys, but an aspect ofquality could be observed in theenergy efficiency and cost of heat-ing and cooling of their homes.Households with low incomes hadhigher costs associated with heat-ing and cooling on a per-square-
foot and per-degree-day basis;these costs fell greatly over theperiod 1981 to 2001, and to aboutthe same extent as for otherhouseholds.
Lower-income households gainedground on higher-income house-holds in the possession of con-sumer durables. Over the periodfrom 1992 to 2002, the ownershipof most consumer durablesbecame more equally distributed
across expenditure deciles, whichmeant that the share going to thelowest-expenditure householdsincreased. The association ofmaterial well-being measures withpoverty over this period was notexamined.
Further evidence on areas of mate-rial well-being other than consumerdurables from SIPP data collectedin 2003 will be available later.
U.S. Census Bureau 19
Table 10.Home Space Cooling Consumption and Expenditures by Selected HouseholdCharacteristics: 1981 and 2001
Characteristic
1981 2001
Dollars for airconditioning
per household(inflation-adjusted
U.S. dollars)
Cooling Btu1
intensity(Btus perthousand
square feetand
degree-day2)
Cooling dollarintensity
(inflation-adjusted U.S.
cents perthousand
square feet anddegree-day2)
Dollars for airconditioning
per household(inflation-adjusted
U.S. dollars)
Cooling Btu1
intensity(Btus perthousand
square feetand
degree-day2)
Cooling dollarintensity
(inflation-adjusted U.S.
cents perthousand
square feet anddegree-day2)
All households (that use electricityfor central air-conditioning) . . . . . . . . 231 4.5 14.6 193 2.8 7.1
Household incomeLess than $10,000. . . . . . . . . . . . . . . . . 170 4.3 14.2 126 3.1 7.7$10,000 to $14,999. . . . . . . . . . . . . . . . 190 4.4 14.0 132 3.0 7.3$15,000 to $19,999. . . . . . . . . . . . . . . . 196 4.8 15.2 159 3.2 7.9$20,000 to $29,999. . . . . . . . . . . . . . . . 240 4.5 14.8 143 2.9 7.1$30,000 to $39,999. . . . . . . . . . . . . . . . 263 4.7 14.7 170 3.1 7.8$40,000 to $49,999. . . . . . . . . . . . . . . . 324 4.4 13.8 187 2.8 6.9$50,000 to $74,999. . . . . . . . . . . . . . . . 383 4.7 16.0 230 2.9 7.2$75,000 or more . . . . . . . . . . . . . . . . . . 444 5.3 19.5 280 2.6 6.7
Income relative to poverty thresholdIncome below poverty threshold. . . . . 193 4.8 15.7 142 3.4 8.2Income 1 to 1.25 times poverty . . . . . 161 3.7 13.4 170 3.4 8.1Income above 1.25 times poverty . . . 239 4.5 14.7 202 2.8 7.0
1British thermal unit (Btu): The quantity of heat required to raise the temperature of 1 pound of liquid water by 1 degree Fahrenheit at thetemperature at which water has its greatest density (approximately 39 degrees Fahrenheit).
2Cooling degree-days (CDD): A measure of how warm a location is over a period of time relative to a base temperature, most commonlyspecified as 65 degrees Fahrenheit. The measure is computed for each day by subtracting the average of the day’s high and low tempera-tures from the base temperature (65 degrees), with negative values set equal to zero. Each day’s cooling degree-days are summed to cre-ate a cooling degree-day measure for a specified reference period. Cooling degree-days are used in energy analysis as an indication of air-conditioning energy requirements or use.
Notes: Household income is income from all sources, self-reported by the household respondent. Consumption and expenditures are forelectricity used for central air-conditioning.
Source: Energy Information Administration, Office of Energy Markets and End Use, Forms EIA-457 A-G of the 1981 Residential EnergyConsumption Survey and 2001 Residential Energy Consumption Survey.
References
Agresti, Alan, Categorical DataAnalysis, Second Edition, WileySeries in Probability andStatistics, Number 359, JohnWiley & Sons, New York, 2002.
Bauman, Kurt J., “Age and MaterialWell-Being in the Survey ofIncome and ProgramParticipation,” 2003Proceedings, American StatisticalAssociation, Social StatisticsSection, 2004.
Bauman, Kurt J., ExtendedMeasures of Well-Being: LivingConditions in the United States:1998, Current PopulationReports, P70-87, U.S. CensusBureau, Washington, DC, 2003.
Bauman, Kurt J., “Shifting FamilyDefinitions: The Effect ofCohabitation and OtherNonfamily HouseholdRelationships on Measures ofPoverty,” Demography 36(3):315–325, 1999.
Bresnahan, Timothy F. and RobertJ. Gordan (eds.), The Economicsof New Goods, Studies Incomeand Wealth Volume 58, NationalBureau of Economic Research,The University of Chicago Press,Chicago, IL,1997.
Citro, Constance. F. and Robert T.Michael (eds.), MeasuringPoverty: A New Approach,National Academy Press,Washington, DC, 1995.
Coder, John and Lydia Scoon-Rogers, “Evaluating the Qualityof Income Data Collected in theAnnual Supplement to the March
Current Population Survey andthe Survey of Income andProgram Participation,” Survey ofIncome and ProgramParticipation Working PaperSeries, No. 215, U.S. CensusBureau, Washington, DC, 1996.
DeNavas-Walt, Carmen, Robert W.Cleveland, and Bruce H.Webster, Jr., Income in theUnited States, 2002, CurrentPopulation Reports, P60-221,U.S. Census Bureau, Washington,DC, 2003.
Fienberg, Stephen E., The Analysisof Cross-Classified CategoricalData, Second Edition, MIT Press,Cambridge, MA, 1980.
Greenspan, Alan, “IncomeInequality: Issues and PolicyOptions,” Remarks at a sympo-sium sponsored by the FederalReserve Bank of Kansas City,Jackson Hole, Wyoming,(August), 1998.
Liegey, Paul, “Hedonic QualityAdjustment Methods forMicrowave Ovens in the U.S.CPI,” BLS Draft Paper, Bureau ofLabor Statistics,<www.bls.gov/cpi/cpimwo.htm>accessed June 25, 2004,Washington, DC, 2001.
Liegey, Paul and Nicole Shepler,“Adjusting VCR Prices for QualityChange: A Study Using HedonicMethods,” Monthly Labor Review(September): 22–37, 1999.
Mayer, Susan and ChristopherJencks, “Recent Trends inEconomic Inequality in theUnited States: Income versusExpenditures versus Material
Well-Being,” in D. B.Papadimitriou and E. N. Wolff,Ed., Poverty and Prosperity inthe USA in the Late TwentiethCentury, St. Martin’s Press, NewYork, 1993.
Proctor, Bernadette D. and JosephDalaker, Poverty in the UnitedStates, 2002, Current PopulationReports, P60-222, U.S. CensusBureau, Washington, DC, 2003.
Short, Kathleen S. and MartinaShea, Beyond Poverty, ExtendedMeasures of Well-Being: 1992,Current Population Reports, P70-50RV, U.S. Census Bureau,Washington, DC, 1995.
Slesnick, Daniel, “Gaining Ground:Poverty in the Postwar UnitedStates,” Journal of PoliticalEconomy 101:1, 1993.
U.S. Census Bureau, Money Incomeof Families and Persons in theUnited States: 1978, CurrentPopulation Reports, P60-123,U.S. Census Bureau, Washington,DC, 1980.
U.S. Census Bureau, SupplementalMeasures of Material Well-Being:Expenditures, Consumption andPoverty 1998 and 2001, CurrentPopulation Reports, P23-201,U.S. Census Bureau, Washington,DC, 2003.
U.S. Department of Health andHuman Services, Measures ofMaterial Hardship, U.S.Department of Health andHuman Services, Office of theAssistant Secretary for Planningand Evaluation, (April),Washington, DC, 2004.
20 U.S. Census Bureau
The Survey of Income andProgram Participation
The data in Tables 1 to 3 and inFigures 1 and 2 in this report werecollected from three panels of theSurvey of Income and ProgramParticipation (SIPP). Data for 1992were collected from two panels:the 1991 panel third wave (inter-view) and the 1992 panel sixthwave. Data from both these panelswere collected in the same months:September to December of 1992.These data do not include imputa-tions for item nonresponse, there-fore frequencies are based onactual responses. For the mostpart, nonresponse levels for thesequestions were in the range of 1 to2 percent. Data from the 1996panel were collected during Augustthrough November 1998 in theeighth wave. Unlike the data fromthe 1991 and 1992 panels, thesedata include imputations for nonre-sponse. The comparison of num-bers of households in Table 2 isaffected by the change in handlingof missing responses. However,the imputations have little impacton comparisons of percentages.
The SIPP is a longitudinal surveyconducted at four-month intervals.The population represented (thepopulation universe) is the civiliannoninstitutionalized population liv-ing in the United States. The insti-tutionalized population, which isexcluded from the population uni-verse, is composed primarily of thepopulation in correctional institu-tions and nursing homes (91 per-cent of the institutionalized popula-tion [4.1 million] in Census 2000).
Statistics from sample surveys aresubject to sampling and nonsam-pling error. All comparisons pre-sented in the sections of the reportusing SIPP data have taken sam-
pling error into account and aresignificant at the 90-percent confi-dence level. This means the 90-percent confidence interval for thedifference between the estimatesbeing compared does not includezero. Nonsampling errors in sur-veys may be attributed to a varietyof sources, such as how the surveywas designed, how respondentsinterpret questions, how able andwilling respondents are to providecorrect answers, and how accu-rately answers are coded and clas-sified. The Census Bureau employsquality control proceduresthroughout the productionprocess, including the overalldesign of surveys, testing thewording of questions, reviewingthe work of interviewers andcoders, and conducting statisticalreview of reports, to minimize thechance of errors.
The SIPP employs ratio estimation,whereby sample estimates areadjusted to independent estimatesof the national population by age,race, sex, and Hispanic origin. Thisweighting procedure partially cor-rects for bias because of undercov-erage, but how it affects differentvariables in the survey is not pre-cisely known. Moreover, biases mayalso be present when people whoare missed in the survey differ fromthose interviewed in ways otherthan the categories used in weight-ing (age, race, sex, and Hispanicorigin). All of these considerationsaffect comparisons across differentsurveys or data sources.
Information on the source of dataand the accuracy of estimates fromthe 1991 and 1992 panels of theSIPP, including the use and compu-tation of standard errors, is avail-able in the “Source and AccuracyStatement for the 1992 Public UseFiles From the Survey of Income
and Program Participation,” at<www.sipp.census.gov/sipp/sourceac/S&A92_puf.pdf>.Information on the source of dataand the accuracy of estimates fromthe 1996 panel is available in the“Source and Accuracy Statementfor the 1996 Public Use Files Fromthe Survey of Income and ProgramParticipation,” at <www.sipp.census.gov/sipp/sourceac/s&a96_040501.pdf>. For furtherinformation on statistical standardsand the computation and use ofstandard errors, contact John L.Boies, U.S. Census Bureau,Demographic Statistical MethodsDivision, at 301-763-4150 or viaInternet e-mail ([email protected]).
The Consumer Expenditure Survey
The current Consumer Expenditure(CE) Survey program began in1980. The Census Bureau con-ducts the survey for the Bureau ofLabor Statistics. The principalobjective of the survey is to collectinformation on the buying habitsof consumers living in the UnitedStates. The survey consists of twocomponents:
■ A Diary, or recordkeeping, sur-vey completed by participatingconsumer units for two consecu-tive one-week periods.27
■ An Interview survey in whichexpenditures of consumer unitsare obtained in five interviewsconducted every three months;the inventory information ana-lyzed in this study are from theInterview survey.
Survey participants report dollaramounts for goods and services
U.S. Census Bureau 21
APPENDIX A: DATA SOURCES
27 For a definition of consumer units, seefootnote 19.
purchased during the reportingperiod, regardless of whether pay-ment is made at the time of pur-chase. Expenditure amountsinclude all sales and excise taxesfor all items purchased by the con-sumer unit for itself or for others.Excluded from both surveys are allbusiness-related expenditures andexpenditures for which the con-sumer unit is reimbursed.
Each component of the surveyqueries an independent sample ofconsumer units that is representa-tive of the U.S. population. In theDiary survey, about 7,500 consumerunits are sampled each year. Eachconsumer unit keeps a diary for twoone-week periods, yielding approxi-mately 15,000 diaries a year. TheInterview sample is selected on arotating-panel basis, surveyingabout 7,500 consumer units eachquarter. Each consumer unit isinterviewed once per quarter, forfive consecutive quarters. Data arecollected on an ongoing basis in105 areas of the United States.
The Interview survey is designed tocapture expenditure data thatrespondents can reasonably recallfor a period of three months orlonger. In general, the data cap-tured are relatively large expendi-tures, such as spending on realproperty, automobiles, and majorappliances, or expenditures thatoccur on a regular basis, such asspending on rent, utilities, andinsurance premiums. Includingglobal estimates of spending forfood, it is estimated that about 95percent of expenditures are coveredin the Interview survey.Expenditures on nonprescriptiondrugs, household supplies, and per-sonal care items are excluded, butare collected in the Diary survey.The Interview survey also providesdata on expenditures incurred onleisure trips. The Diary survey isdesigned to capture expenditures
on small, frequently purchaseditems that are normally difficult forrespondents to recall. Detailedrecords of expenses are kept forfood and beverages (both at homeand in eating places) tobacco,housekeeping supplies, nonpre-scription drugs, and personal careproducts and services. Expendituresincurred away from home overnightor longer are excluded from theDiary survey. Although the diarywas designed to collect informationon expenditures that could not berecalled easily over a given period,respondents are asked to report allexpenses (except overnight travelexpenses) that the consumer unitincurs during the survey week.
The Residential EnergyConsumption Survey
The Residential Energy ConsumptionSurvey (RECS) is the EnergyInformation Administration’s (EIA)benchmark national survey provid-ing data on energy consumptionand expenditures in conjunctionwith characteristics of housing unitsand their residents. The RECS isconducted every 4 years, mostrecently for data year 2001, theeleventh time the survey was con-ducted. Data are collected via vol-untary computer-assisted personalinterviews with a probability sampleof about 5,000 housing units nation-wide, and via mandatory follow-upmail collection of energy data fromthe sample households’ energy sup-pliers. Almost all of the housingunit data are provided by a responsi-ble householder, but the interviewerdoes measure the floor space of thehousing unit, which is a crucial vari-able explaining energy use.
The RECS estimates are based on astatistical sample using an areaprobability sampling design. Thesampling unit is the housing unit.The information obtained in theRECS is used to construct adatabase on the household sector
describing the consumption anduse of energy and the characteris-tics of the consumers. Publicationsand electronic data files can beviewed and downloaded from theInternet through the EIA Home Page<www.eia.doe.gov> or more directlyat <www.eia.doe.gov/emeu/recs>.
The survey is conducted in twophases. In the first phase, theenergy-related characteristics ofthe household and data on thefuels and equipment used in thehousing unit are collected during apersonal interview with an eligibleadult member of the household.Building and energy characteristicsfor rental units are collected fromrental agents, landlords, and apart-ment managers during telephoneinterviews. The second phase ofRECS collects data on fuel con-sumption and expenditures byhousing unit from the suppliers ofenergy to the housing units in theRECS sample.
The RECS also collects data for theU.S. Department of Health andHuman Services (HHS)Administration for Children andFamilies (ACF). In 2001, ACFfunded a supplemental survey of500 Low Income Housing EnergyAssistance Program (LIHEAP) recipi-ent households residing in theareas targeted for sampling. Priorto 2001, the RECS collected LIHEAPdata from a supplemental sampleof approximately 800 low-incomehouseholds.
Among the household characteristicscollected are the number of house-hold members and total householdincome from all sources. These datapermit the construction of identifiersof household income relative to thepoverty line and an identifier for eli-gibility for federal assistance (i.e.,LIHEAP). The former identifiers arethe basis for the tables and figurespresented in this report.
22 U.S. Census Bureau
The Concentration Index is com-puted as one minus twice the areaunder the concentration curve(defined below). This is similar tothe equation used to compute theGini coefficient; however, the Giniis based on the area under theLorenz curve.28 The curves repre-sent cumulative probability distri-butions (CPD).
Consider a variable, y, that is somefunction of total expenditures, x;i.e., y = g(x), and consumer unitsare ranked in ascending order ofx—their total expenditures in thiscase. The concentration curve fory is defined as the share of total y,for example the total percentage ofconsumer units who own a televi-sion, for consumer units with totalexpenditures of x or less, Fi[g(x)],graphed against the populationshare of those with total expendi-tures no greater than x, F(x). The
concentration index is one minustwice the area under the concentra-tion curve; i.e.,
(1)
The concentration index liesbetween –1 and +1. It could benegative if low-income people hada higher ownership rate than high-income people. This is in contrastto the Gini coefficient that liesbetween zero and one.
Using percent ownership
To produce the concentrationindexes for ownership, the percentownership rates by decile aretransformed into a discrete proba-bility distribution and replace y inequation 1. This is shown in equa-tion 2:
(2)
where pji is the fraction of allhouseholds that own durable goodj that are in total expendituredecile i and rji is the actual owner-ship rate of good j for the ithdecile. By construction, the sum ofthe pjis is equal to one. For goods
that have ownership rates that arerelatively equal across deciles,regardless of the level of the own-ership rate, the probability distri-butions are fairly flat with valuesfor pji close to 0.1. For goods thatare more concentrated among theaffluent households, the probabil-ity distributions tend to rise acrossthe income deciles.
Using average number owned
When using the concentrationindex to describe the distributionof average number of durablegoods owned, the computation isthe same as in equation 2 exceptthat the percent ownership in eachdecile is replaced by the averagenumber owned in each decile.
Using average value
For each decile, the average expen-diture on durable goods for thoseconsumer units that purchased adurable good is calculated. Thisamount is multiplied by the aver-age number of vehicles owned (bydecile). This yields a proxy for thevalue of the vehicles owned byeach decile.
∑ ==
10
1i jijiji rrp
x
Cy = 1 - 2 I Fi [ g (x) ] d F ( x )0
U.S. Census Bureau 23
APPENDIX B: CONCENTRATION INDEXES FOR DURABLE GOODS
28 The Gini coefficient (or index of incomeconcentration) is a statistical measure ofincome equality ranging from 0 to 1. Ameasure of 1 indicates perfect inequality;i.e., one person has all the income and therest have none. A measure of 0 indicatesperfect equality; i.e. all people having equalshares of income. For a more detailed dis-cussion, see U.S. Census Bureau, 1980.
This appendix discusses severalapproaches that can be taken tocompare the relative performanceof households with income aboveand below the poverty threshold,with a focus on the example ofcomputer ownership by povertystatus in 1992 and 1998.
Relative performance of house-holds is not as easily described asabsolute performance, because rel-ative performance can bedescribed in different ways.Percentage point change in thepercent of households having aparticular material well-being indi-cator is the percent having an indi-cator in 1998 minus the percenthaving that indicator in 1992. Theprevalence of computers in house-holds with income at or abovepoverty grew from 23 percent to45 percent from 1992 to 1998,while the corresponding growth inhouseholds with income belowpoverty was from 7 percent to 18percent (Table 11). The percent-age-point shift in possession ofcomputers is the 1998 percentageminus the 1992 percentage, or 23percentage points for householdsabove poverty and 11 percentagepoints for households belowpoverty (see the first two columnsof Table 12).
Another common measure of com-parative growth is the percentchange in the proportion of house-holds having a material well-beingindicator. Using the same exam-ple, the proportion of householdswith income at or above povertypossessing computers doubledduring the 1992 to 1998 period(23 percent to 45 percent), for a100 percent gain. Over the same
period, households with incomebelow poverty increased their rateof computer ownership by 2.6times (7 percent to 18 percent), fora 161 percent gain.
The two methods for examiningthe relative increase in computerownership by households withincome at or above poverty andhouseholds with income belowpoverty yield opposite conclusions.The 23 percentage-point gain incomputer ownership by house-holds at or above the poverty levelwas larger than the 11 percentage-point gain of households belowpoverty. At the same time, the100 percent gain in proportionsowning a computer of householdsat or above the poverty level wassmaller than the 161 percent gainof households below poverty.
To address these complications,statisticians have proposed theodds ratio as a possible way torepresent these relationships(Fienberg, 1980; Agresti, 2002).“Odds” compare the probabilitythat an event will occur to theprobability it will not, and the“odds ratio” is the ratio of twoodds. The odds of a householdwith income at or above thepoverty level having a computer in1992 was the percentage that hada computer (23 percent) divided bythe percentage that did not have acomputer (77 percent), or 0.29.The odds of a household withincome below the poverty levelhaving a computer in 1992 was thepercentage that had a computer (7percent) divided by the percentagethat did not have a computer (93percent) or 0.08. The odds ratiofavoring computer ownership
among households with income ator above poverty in 1992 was 0.29divided by 0.08, or 3.9. In otherwords, households with income ator above poverty had almost 4times the odds of having a com-puter as households with incomebelow poverty.
The odds ratio of 3.9 showsgreater prevalence of computersamong households with income ator above poverty than amonghouseholds with income belowpoverty, and confirms what wasstated above using simpler statis-tics: households with income at orabove poverty were more likely toown computers. Where the oddsratios provide additional insight isin the examination of changesbetween 1992 and 1998. In 1998the odds ratio favoring computerownership among households withincome at or above poverty was3.7, not statistically different fromthe value of 3.9 observed in 1992.In other words, as measured by theodds ratio, the relative computerownership by households withincome at or above poverty andhouseholds with income belowpoverty did not change between1992 and 1998.
Across all 12 items on which odds-ratio comparisons between 1992and 1998 were made (Table 12),none showed a statistically signifi-cant change. Despite the absoluteincrease in 8 of these materialwell-being indicators by house-holds with income below poverty,there was no increase or decreasein their position relative to house-holds at or above poverty duringthis period.
24 U.S. Census Bureau
APPENDIX C: STATISTICAL APPROACHES TO COMPARING TRENDS IN MATERIAL WELL-BEING
Tables with additional informationon the topics covered in this reportare available from each of thethree contributing agencies. TheCensus Bureau has produced sev-eral reports on measures of mate-rial well-being, including detailedtables, which are available at
<www.census.gov/population/www/socdemo/wellbeing.html>.
Additional tables on possession ofconsumer durables and relatedinformation from the ConsumerExpenditure Survey are availablefrom the Bureau of Labor Statisticsat <www.bls.gov/cex/home.htm>.
Also, the Energy InformationAdministration has produced a setof detailed tables on energy use byhouseholds, energy efficiency ofheating and cooling, and other top-ics, at <www.eia.doe.gov/emeu/recs/contents.html>.
U.S. Census Bureau 25
Table 12.Comparison of Indicators of Material Well-Being of Households by Poverty Status UsingThree Measures—Percentage Point Change, Percent Change in Proportions, and OddsRatio: 1992 and 1998
Percentage point change inhouseholds, 1992 to 1998
Percentage change inproportion of households,
1992 to 1998
Odds ratio of householdsabove poverty and
below poverty1
At or abovepoverty
Belowpoverty
At or abovepoverty
Belowpoverty 1992 1998
Stove . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . –0.3 –0.4 -0.3 –0.4 3.9 3.2Enough of food wanted . . . . . . . . . . . . . . . . . –0.1 1.3 -0.1 1.4 7.0 5.5Telephone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.7 5.9 0.8 7.3 6.9 5.7No smoke or odors in neighborhood . . . . . . 2.3 3.6 2.4 4.0 1.6 1.5No unpaid rent or mortgage . . . . . . . . . . . . . 1.9 5.4 2.0 6.7 3.5 3.4No unmet need for doctor . . . . . . . . . . . . . . . 1.5 1.7 1.6 2.1 2.7 3.1No roof or ceiling leaks . . . . . . . . . . . . . . . . . 1.5 3.6 1.7 4.2 2.0 1.8No abandoned houses in neighborhood . . . 2.0 5.1 2.2 6.1 1.9 1.6No unmet essential expenses. . . . . . . . . . . . –0.1 0.9 -0.1 1.3 3.5 3.3Air conditioner . . . . . . . . . . . . . . . . . . . . . . . . . 8.0 15.1 11.2 28.7 2.2 1.8Dishwasher . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.9 5.7 12.9 28.6 4.6 4.4Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.7 11.4 100.0 161.2 3.9 3.7
1 For definition of odds ratio, see text.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
ADDITIONAL DATA
Table 11.Percentage of Households With Selected Measures of Material Well-Being by PovertyStatus: 1992 and 1998
IndicatorAll households At or above poverty Below poverty
1992 1998 1992 1998 1992 1998
Stove . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98.9 98.7 99.2 99.0 97.1 96.7Enough of food wanted . . . . . . . . . . . . . . . . . . . . 97.6 97.8 98.6 98.5 91.1 92.4Telephone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94.7 96.2 96.7 97.5 81.1 87.0No smoke or odors in neighborhood . . . . . . . . . 92.6 95.1 93.1 95.4 89.7 93.2No unpaid rent or mortgage . . . . . . . . . . . . . . . . 92.2 94.6 93.8 95.7 81.4 86.8No unmet need for doctor . . . . . . . . . . . . . . . . . . 92.3 93.9 93.4 95.0 84.2 85.9No roof or ceiling leaks . . . . . . . . . . . . . . . . . . . . 91.4 93.1 92.1 93.7 85.5 89.1No abandoned houses in neighborhood. . . . . . 89.6 92.0 90.6 92.6 83.3 88.3No unmet essential expenses. . . . . . . . . . . . . . . 85.9 86.0 88.4 88.3 68.4 69.3Air conditioner . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68.7 77.7 71.1 79.1 52.6 67.7Dishwasher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49.3 56.0 53.4 60.3 20.0 25.7Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.7 42.0 22.6 45.3 7.1 18.4
Source: U.S. Census Bureau, Survey of Income and Program Participation, 1991, 1992, and 1996 panels.
U.S. Department of CommerceEconomics and Statistics AdministrationU.S. CENSUS BUREAU
Washington, DC 20233
OFFICIAL BUSINESS
Penalty for Private Use $300
FIRST-CLASS MAILPOSTAGE & FEES PAIDU.S. Census Bureau
Permit No. G-58