(Archived document, may contain errors)
January 31,1992 HOW THE POOR REALLY LIVE LESSONS FOR WEL,FARE IUWORM INTRODUCTION Each yecli for the past ~rr?~~a the united states Census ~llieau has npar~d that at least 30 million Americans live in poverty. For odmy Americans, the word"p0ver ty" suggests destitutiun, and the idea ofwbspmd and unremitting poverty in aland ofprospuity is of course deeply dismssing.The visiun is one of over 10 percent ofthe entire populatian being hungry and malnourish ed, living in over-wwdcd, poorly heated, filthy apartments.
Yet, this simply is not the case.
Sophisbtd surveys ofthe actual living conditions of America's poor by the Cen sus Bureau, the U.S. DeparmDent of Agricultun, and other government agencies, paint a vuy different pictm.These smys show that the overwhelming majority of per sons ofkiallyidcntifiedas"poar,"min fact well fedand well housed.
That has major implications farthe design of American welfm policy. A distorted picture ofthe conditions af low- income families has led to misdirected policies that, in too many instances, give the wrong help to the wrong people while ignoxing the real needs of many lower incame Americans drive federal policy is that policy makers, curiously, are not looking at the federal government's own data. As an earlier Heritage Foundation study indicated, for in stance+ because lawmakers and federal o&cials ignore crucial data on nutrition fiom the U.S. Deparrment of Agriculture and the U.S. Centers far Disease Control, many b elieve there is widespread hunger in America. In fact, there is almost no poverty-in dud malnutrition in America, yd the poor have virtually the same level of nutrient consumption as the middle class Crucial Data Ignored. The mason that an inaccurate picm of poverty continues to 1 Robar Recux Food F
t Policy Review, No. 58, Fall 1991, pp. 3843; Robert Rector Hunger and Malnutrition mong America's chilbna Herimp Foundation BucLgrorrndnNo. 843, August 2,1991.
A distorted picture of the housing conditions of Americans classified as poor by the Census Bureau is also prevalent among policy makers. Again, readily available data refute many of the impressions driving policy. For example; among the little-known fact s about poor housing from the 1989 American Housing Survey, a joint project of the U.S. B au of the Census and the Department of Housing and Urban Develop ment are F Nearly 40 percent of all poor households actually own their own homes? The median vafue of the homes of these households is 58 per cent of the median value of all homes owned in America. Over one mil lion poor households owned homes worth over $80,000 The average home Owned by persons classified as poor in the U.S. is a three-bedroom house with a garage and porch or patio. Contrary to popular impression, the majority of these households who own their own homes are not elderly According to the Census Bureau, only 8 percent of poor households are overcrowded. Nearly two-thirds have more than two r o oms per per son The average American defined as poor has twice as much living space as the average Japanese and four times as much living space as the average Russian. Note: These comparisons are to the average citizens in these countries, not to those cl assified as poor The homes and apartments of the poor typically are in good condition.
The Census Bureau reports that only 5 percent of all housing of the poor have even moderate upkeep problems Some 53 percent of poor households, owners as well as renters , have air conditioning. By contrast, just twenty years ago only 36 percent of the entire U.S. population enjoyed air conditioning Contrary to popular impression, housing costs for many poor households are quite low; half of all poor households either liv e in tax payer-subsidized public housing or own their own homes with mortgages fully paid 2 U.S. Department of Commerce, Bureau of the Census, and U.S. Department of Housing and Urban Development Wrce of Policy Development and Resmh, American Housing Surve y for rhe United Srutes in 1989, Current Housing Reports H150/89 Washington, D.C U.S. Government Printing Office, July 1991) Unless otherwise noted all dam in this paper m.hm the 1989 American Housing Survey 3 Henceforth, the word poor will refer to those i ndividuals or households classified as such by the Census Bureau 4 See definition in text, p. 8 5 For purposes of simplicity, the term public housing will be used in this paper to refer to all federal, state, and local subsidized rental housing. This incl udes HUD public housing, HUD Section 8 housing, other HUD rental subsidy programs, as well as other state and local public housing 2 Overall, the actual housing and living standards of the poor are far higher than the public would imagine.
Much public misu nderstanding about poverty occurs because the Census Bureaus annual poverty report greatly overstates the extent of poverty in America. The Census Bureau defines as poor any household with a cash income less than the officially specified poverty income th r eshold In 1990, the latest year for which the calculation has been made, the poverty income threshold for a family of four was 13,359 There are two key flaws in the federal governments measurement First, when calculating a familys financial situation, the Census Bureau ignores all of the familys existing assets. Cars, homes, and other valuables have no bearing on the determination. When private sector financial analysts determine a familys finan cial condition, of course, such assets are included. The Cens u s Bureau counts only in come acquired during the year. Under these limited criteria, a small businessman who loses money during the current year is defined as poor, even though he may own as sets worth hundreds of thousands of dollars Second, the Census B u reau also greatly understates the current incomes of lower-in come families For one thing, the Bureau misses cash earnings that families do not report to tax collectors; for another, it ignores 158 billion in welfare benefits that go to these families. Th e benefits ignored by the Bureau include much cash welfare, food stamps, Medicaid, public housing subsidies and most other welfare spending. The total missing welfare spending comes to an average of $1 1,120 for every poor household in America.
To give acc urate information to members of Congress and federal policy makers about poverty so that they can design better welfare programs, the federal government should revise its measurement of poverty. The current Census poverty report lacks credibility because i t does not provide complete and accurate infarmation on the living conditions of Americas poor. It should be replaced with a new government assess ment of poverty. To obtain this, the government should 1) Integrate and expand existing government surveys m e asuring the actual physical living conditions of low income families. This would include meas urements of food consumption, nutrition, housing conditions, and property ownership 2) Create a new measure of financial resources available to low income per so ns. The Census Bureaus official measure of income now is very inaccurate.
Other surveys conducted by the Bureau, for example, reveal that low-income households in 1989 spent $1.94 for every $1.00 of income reported by the Census Bureau6 Instead of publishi ng flawed income surveys the federal government should present Congress and the American people with an ac curate assessment of the financial status of low-income households based on 6 See Robert Rector, How Poor Are Americas Poor? Heritage Foundation Bac kgrounder No. 791, September 21 1990 3 actual household expenditures and assets, as well as a correct count of the bil lions of dollars in welfare benefits received each year THE LIVING CONDITIONS OF THE POOR Note: All figures am in percent.
Data from 1985 Amencan Housing Sutvey; later figures are not available Havin all toilets stopped u lwo or more times in the last 3 months. 80urce:k.S. Departmen! of 8ommerce, Bureau of the Census, and U.S. Department of Housin and Urban .Devolpment, otfice of Poll Deve lo men! Research, Amencan Housm Suyey for rhe Unrted8ates m 19
89. Cumnt Housing Reports &0/89 (hashingpn. D.C US. Governmen! Pnnsng Office, July 1991 The family income data of the Census Bureau leave the impression among most Americans that millions of Am erican households are malnourished and lack adequate housing. According to the 1991 Census Bureau report, there were 33 million poor per sons in the U.S. in 1990.7 But few of-the 33 million persons defined as poor actually fit the public image of poverty. Scientific surveys conducted by the U.S. Department of Agriculture, U.S. Centers for Disease Control, and even the U.S. Census Bureau it self, show that the actual living conditions of pooryy persons are far better than most Americans imagine Table 1 Kous i na Conditions: Poor Households Free from all external defects broken windows, sagging roof, missing roof material, sloplng walls, and crack, in foundation 88.9% 91 .o% 87.5 I 97.4 98.7 96.6 I Has full kltchen I (stove. oven, slnk. and refrigerator I Free f rom recurrent tollel I stoppages or breaks In prlor three I 98.2 98.9 97.6 7 U.S. Department of Commerce, Bureau of the Census, Measuring the Effect of Benefits and Taxes on Income and Poverty: 1990, Series P-60, No. 176-RDl (Washington, D.C U.S. Governme n t Printing Office, August 1991 p 11 4 In reality, for example, there is little or no poverty-induced malnutrition in America poor gersons have virtually the same levels of nutrient consumption as the middle class. Poor children today actually are super-no u rished compared with the children of previous American generations. Today's poor children will grow up to be significantly taller and heavier than the av rage child of the same age in the general American population of the late 1950s. The ovemding dietary problem among low-income American families is not a lack of nutritious food, it is obesity.
Similarly, the popular image of the housing of the poor is of a dilapidated building with broken or boarded-up windows. In reality, very few poor homes fit this de scrip tion. The overwhelming majority of poor persons live in housing that is roomy, struc turally sound, and in good repair.
Table 1 is based on the Census Bureau's "American Housing Survey" of 1989, the 8 latest available data on the housing conditions of poor households. According to the survey, the majority of poor households live in single homes with two or three bedrooms. AsTable 1 shows, some 89 percent of homes and apartments of the poor a~ free of external problems, such as a sagging roof or crac ks in the founda tion; 95 percent are free of major defects such as or boarded-up windows.
And only one out of ten homes housing the poor suf fers from internal main tenance problems such as peeling paint, cracked plaster on interior walls signs of rodents in the last three months, or at least one heating problem during the previous winter 9P Chart 1 Poor 'Households: Owner vs. Renters 1989 Elderly Owner on-Elderly Owner Yortagr Pald In Full Mortagr Pald In Full Non-Subaldlzad Hourlng bouroeI 1989 Amerlcen Houelng Survey.
Horlte~e D8taChrrt 8 Rector Food Fight op. cit and Rector Hunger and Malnutrition Among.America's Children op. cit 9 Bernard D. Kaxpinos, Height and Weight of Military rouths (Medical Statistics Division, Office of the Surgeon General, Dep artment of the Army, 1960 pp. 336-3
51. Information on the current height and weight of youths pvided by the National Center for Health Statistics of the U.S. Department of Health and Human Services based on the National Health and Nutrition Examination S urvey 10 The estimate of the percent of housing units free from all external defects is based on those units in which a complete examination of the external conditions of the house was made by the Census Bureau. Housing units for which complete observatio n s were not made were excluded from the calculation. American Housing Survey 1985 HOME 0 DecentQuality. The vast majority ofpoor households live in housing free of even minor problems. The quality of housing for poor households, in short, is far superior t o the image of poverty conjured up in the minds of most Americans.
Another surprising fact is that 39.5 percent of all households classified as poor by the Census Bureau actually own their own homes.The average home owned by the poor, moreover, is a three-bedroom house in a non-urban area. And as Chart 1 shows close to t h ree-quarters of home-owning poor families have fully paid off all mortgages on their homes. While the majority of poor households with fully paid mortgages are elderly, there are nearly 1.5 million non-elderly, poor households who own homes free of mortga ges. These households constitute 11 percent of all poor households in the U.S Chart 1 also gives some indication of the cost of housing for low-income households. It shows that 22 percent of all poor households reside in public housing.
This means that 61. 5 percent of poor households either own their own homes or live in heavily subsidized public housing-in either case, housing payments are very low, or even zero NERSHIP AMONG THE POOR Contrary to common per ceptions, most of the 40 per cent of poor househ o lds owning their own homes are not elderly, as Chart 2 shows. Families with children comprise one-third of the home-owning v another onequarter are non elderly households without children. Elderly households without children make up the remaining 45 perce nt.
While they m not a majority of the home owning poor, elderly poor householders BT~ more like ly to own homes than are other types of poor households. As Chart 3 shows, over 60 percent of all poor, elderly households own homes. By comparison the rate of home ownership among poor families with Chart 2 Poor Owner Households 1989 Non-Married With Children Non-Elderly Elderly Owroo8 1080 Amerlcen HouelnO Survey. Gr Heritape DatrChrrl children is less than half the rate of the elderly-or 27 percent. And roug hlyone-third of those poor families who are childless and not elderly own their homes 6 To be sure, Americans classed as poor do not live in villas or mansions. But neither do many live in slums.
The average'house owned by poor households is of reasonable size and in good physical condition. Accord- ing to the data, the typical house owned and occupied by a poor American household was built in the mid-1950s and is eight years older than a home owned and occupied by the'average American family The average p o or home owner has five or six rooms plus bathrooms The house has three bedrooms and one bath and is situated on rough ly a one-half am lot. Over half of homes owned by the poor have a garage or carport Chart 3 Share 'of Poor Households.Owning Home By Fami l y Type All Poor Hourrholdr Eldrrly Hourrholdr Couplo Wlth Chlldron YWrl8d N ohEldor ly WI I houl Chlldnn All Frmllloa Wlth Chlldrmn lllnglr Prnntr Wlth Chlldrrn I I I 1 I 0% 26% 60 76% 100 BoUrOI: 1080 Amerlccln Houslng Survey Horltrgr DrtrChclrt and thre e -quarters have a parch deck or patio tween homes owned by America's poor and those owned by the rest of the American population. Compared with the general U.S. owner-occupied housing stock, homes owned by the poor have one less room and 25 percent less li v ing space. And while nearly two-thirds of all American owner-occupied homes have more than one bathroom, only one third of poor-owner homes have more than a single bath Comparable Homes. These are.modest homes, but there are not vast differences be Low-in come Americans often are the target of derogatory comments about their sup posed unwillingness or inability to keep up their homes. The facts tell a different story.
Homes owned by poor households show a surprisingly high level of upkeep. Accord ing to the latest available data, from the 1985 American Housing Survey, only 2 per cent of poor-owner homes had broken or boarded-up windows; a total of 9 percent had 11 The Census Bureau includes in its count of Moms within a home: bedrooms, living rooms, kitchen s , separate dining rooms, recreation rooms, and other rooms used for year-round living. Bathrooms, closets, porches, garages and unfinished basements are not counted as separate rooms. The Census Bureau slightly undercounts the number of rooms in American h omes and apartments because it counts a dining mom as a separate Mom only if it is separated from all adjoining rooms by walls. Dining moms which open directly into living moms are not counted as separate rooms even though this is a common feature in many homes and apartments. Thus many Moms which would be conventionally thought of as separate dining morns are in fact excluded from the Census count. 9 7 external deficiencies such as broken windows, missing roof material, sloping walls or cracks in the foun c iations.12 ing to the 1989 housing survey. The Census Bureau found only 8 percent had even minor levels of peeling paint or chipped plaster. Overall, just 3.5 percent of poor owner homes were found to have moderate upkeep problems, which the Census Bureau defines as having any three of the following six conditions The interior of nearly all of the homes owned by the poor was in good repair, accord broken plaster or peeling paint on inside walls or ceilings, with at least one area of broken plaster or peeli n g paint as large as a piece of paper eight by eleven inches. signs of mice or rats in the last 90 days Table 2 Facts About Owner-Occupied Housing I I Medlanvearbullt I 1964 1956 I I I Medlan number of auare feet mr Derson I 688 71 3 I I I I I Share wlth a l r condltlonlna I 73.2 I I I I Share wlth three or more bedrooms I 72.1% 58.3% I I I I I Share In central cltles I 24.1% 24.3% I I Source: U.S. Department of Commerce, Bureau of the Census, and U.S. Deptment of Housin and Urban Devel menr, Office of Policy Dewlogent Research, Amerrcan Hwsrng Sme for he 8niW Current Housing Reports Hi 189 (Washington, D.C U.S. GovernmenrPrinting Office Julv 1QQlL I states in 12 Data on external deficiencies are not available for single unit dwellings after 1985 8any crack in interior walls wide enough to allow one to insert a dime; any water leaks in last 90 days from interior pipes or plumbing fixtures 4 any water leaks in last 90 days from roof, windows, or basement any dents or holes in floors cracks need not go through wa ll.
Following the Census Bureaus very specific standards, only eight out of one thousand homes owned by the poor (0.8 percent) were found to have severe upkeep problems, defined as having five out of six of the above deficiencies.
The most common problem affecting poor home owners, in fact, is water leaks from the basement, windows, or roof. Over one out of five poor home owners reported such water leaks during a twelve-month period. Yet this is a problem affecting a wide range of American home owners, no t just poor home owners. In fact, 18 percent of non-poor American home owners reported such water leaks in 1989, a rate virtually identical to that of poor American home owners greater age, homes owned by the poor have a lower retail value than those owned by the general population. The median value of homes owned by the poor was $43,562 in 1989, an amount equal to 58 percent of the median value of all owner-occupied hous ing in America, which was $75,359 in that year. The average home owned by poor per son s is comfortable but modest 13 Poor Households, Expensive Homes. In keeping. with their smaller size and Yet, the 1989 Census survey also showed that about 1.1 million poor households owned homes with values greater than the median value of all U.S. owner- o ccupied housing, both poor and non-poor. Such poor households owning relatively expen sive homes make up over 8 percent of all poor households. The Census Bureau even reports over 215,000 POCK households owned homes worth ov r $200,000, and 75,000 poor ho useholds owned homes worth over $350,000.
The fact that nearly one-tenth of all poor households own homes worth more than the average American home dramatizes how absurd it is for the Census Bureau to ig nore assets in determining whether or not a family i s poor 18 RENTAL HOUSING AMONG THE POOR Over 60 percent of the poor, or 7.5 million households, rent houses or apartments.
While the elderly are the largest group of poor home owners, single parents with children are the largest group of poor renters. Ove r 80 percent of single parents clas sified as poor rent homes, and single parents account for over one-third of all poor households who rent. As Chart 4 shows, members of non-elderly households without children are the next largest group of poor renters, c omprising over one-quarter of the 13 American Housing Survey, 1989, op. cit p. 98 14 Ibid p. 116 9 I total. Married couples with children are the smallest group among poor renters accounting for only 15 per cent of the total The typical poor rental unit w a s built in the mid 1950s and is about six years older than units rented by the non-poor. Nearly 40 per cent of all poor rental hous ing units are single unit dwellings; the remainder are multi-unit apartment buildings, generally with less than twenty apar t ments per building. Over half of all rental units occupied by the poor have a balcony patio, or porch. Some 49 percent have air condition ing. apartment buildings repurt the conventio-nal signs of Few poor renters living in I Chart 4 Poor Renter Household s 1989 5.G I Elderly Married Non-Elderly L Slngle Parents With Children Soum 1989 Amerlcen Houelng Survey.
Horltoao DmtmChrrt slum 1i~ing.l~ Onl; 1 percent report a lack of working lights in building hallways, and 4 percent say their apartment building has broken or boarded-up windows. In apart ment buildings with elevators, 7 percent have elevators that are broken Poor renters a~ nearly twice as likely as poor owners to have problems keeping up their residences; roughly 1 percent of poor renters have seve r e upkeep problems based on the criteria used by the Census Bureau.16 Another 6 percent have moderate upkeep problems, as defined by the Census Bureau.17 In particular, some 11 percent of poor renters report signs of mice or rats during the previous three m onths. One in ten poor renters report that their toilets were broken or plugged for mm than six hours on one or more occasions during a three-month period-but only about 2 percent of poor rental households report frequent toilet breakdowns. The overwhelmi ng majority of poor renters live in housing that is structurally sound, in good repair, free of ro dents, and has adequate heating and plumbing equipment.
There is one major difference between poor renters and both poor and non-poor home owners: poor rente rs are far more likely to be on welfare than are other 15 Data on external problems in 1989 m available only for multi-unit structures 16 See def~tion in text on p. 8 17 Ibid. G 10 live in public housing. Less than half of all poor-renter households repor t any earned income Misleading Figures. Ad Table 3 Facts about Renter Housina I spending or housing assis- 19 tance contend that. not withstanding the condition of their rental housing, the poor pay too great a share of their incomes for rent thus making t h e purchase of other necessities, such as food, difficult or nearly im possible. For example, the Joint Center for Housing Studies of Harvard Univer sity contends that two thirds of all poor renter households pay over 50 per cent of their incomes in housin g costs.18 The Joint Center proposes building many more expensive public housing units to bring down the cost of housing for the poor.
These figures are very misleading, however, be Medlan number of I bedrooms 21 Share In Single unR I dwelllng 37% I Share with balcony, I porch. deck. or atl lo Share In publlc or government-subsldlzed I 14% I rental houslna Medlan monthly houslng I costs (rent. Insurance. and utilities I Source: U.S. Department of Commerce. Bureau of the Census. and US. I Deparbnent of Hous i n and Urban Development, Office of Polit$ Develo ment Resear$ American HwSi Survey for the United States in IWQ, 8urrent Housin heports H150/8Q%ashington, D.C U.S. Government Pnnting &ice July 1QQl cause they rely on the Census Bureau household income fig u res, which greatly under estimate the incomes of poor households. In fact, total spending by these households greatly exceeds their reported incomes, according to the federal governments own figures. Because of this, and since many in-kind benefits availa b le to those families are not counted as income, it should not be surprising to learn that more than 50 percent of their income goes to housing. In order to determine if poor households really are spending too much of their incomes on housing costs, there w ould need to be chan ges in the methodology of the Census Bureau surveys, to present an accurate count of the earnings and financial assets of poor renter households 18 Joint Center for Housing Studies of Haward University, The State of the Nations Housin g 1990 (Cambridge Massachusetts: Joint Center for Housing Studies of Harvard University, 1990) p. 21 11 THE LIVING SPACE OF POOR HOUSEHOLDS Poor Black I 49.0 42.0 9.0 1 Note: All fi ures am In percent Source: U.
8. De artment of Commerce, Bureau of the Cen sus, and U.S. Department of Housin and Urban Develo ment, de of Poli Develo ment Research, American Housi Survey for he Unitedkates in fm Another indicator of housing quality is the amount of living space available within a house. Living space may be meas ured either by the number of moms per person or the square footage of living space available per individual.
In its housing surveys, the Census Bureau counts the number of moms within each housing unit. Bedrooms, living rooms, kitchens, separate dining rooms, recreation morns, and other moms used for year-round living are included in this count.
Bathrooms, closets, porches and unfinished basements are not. On the basis of this sur vey, the Census Bureau then determines the number of rooms av ailable per person within each household. For example: A three-person family living in a the-bedroom house with a kitchen, living Mom, and separate dining room would be counted as having two rooms per person. Households with less than one room per person are defined as crowded. Thus, a family of five living in a two-bedroom apartment with a living mom and kitchen would be regarded as crowded.
Data from the 1989 housing survey indicate that the average poor household actual ly has roughly the same number of rooms per person as the average American household. As Table 4 shows, 60 percent of poor households have two or more rooms Table 4 Crowding in Poor Households I Poorownera I 73.2 22.1 4.7 I 12 the affluent Swiss, 1.
4. Out si& Europe the average Japanese lives in a home with 20 1.25 moms per person.
Another useful measure of the quality of housing is square feet of living space per capita. As Chart 5 shows poor persons in the U.S. have an average of 405 square feet per capita, or 61 percent of the living space available to the average American. On this the international comparison is striking. While poor Americans have less living space than the average American, they have twice as much living space as the average Japanese and four times the living space of the average citize of the former Soviet Union. 51 Chart 6 Housing Space An International Comparison Square Feet Per Pereon 700 600 TWOO- Tho Living 8DDCO A Awrigr Jrpmnrer Cltlmnr 600 400 800 zoo 100 0 All U.
8. Poor U.
8. All JDDmrDr All Soviet Hourrholdr Hourrholdr Hourrholdr Hourrholdr 8ourorr U.S. DBDErtmBnt of Energy, Energy Information Admlniatratlon, Houelno Chere~ter/sl/cs 1887. 1989.
8. Zaychenb, Unlted States-USSA: lndlvldual Consumption Some Comparlaonal. WrlU Allelrs, 19
89. The Alfluent Japanese Household. Bua/nesa Amerlce. 19
81. Hrrltegr DmtrChmrt WHAT ABOUT CENTRAL CITY HOUSING The term central city poor brings to mind images of rat-infested inner-city tene ments, studded with boarded-up or broken windows. Some of the central city poor do indeed live in squalid housing in dangerous, underclass neighborhoods. But the vast majority of housing for the poor in Americas inner cities is in fact free of the defects commonly associated with slums.
Some 41 percent of American poor househol ds live in central cities. The central city poor are more likely to be single parent families, with family heads less likely to work than are the poor elsewhere in the nation. The average poor central city household lives in a two-bedroom apartment with a balcony, built around 19
48. Fifty percent of poor central city households have air conditioning. And half of the households own a car or truck 19 Living Conditions in OECD Countries: A Compendium of Social Indicators (Paris: Organization for Economic Coq eration and Development. 1986 p. 1
33. European and Japanese figures provide the mean number of rooms per person. Figures for poor American households are also means 24 Ibid..p. 133 21 U.S. Deparunent of Energy, Energy Information Administration, Housing Characteristics 2987 (Washington, D .C Government Printing Office, 1989 p 25. A.S. Zaychenko, United States-USSR Individual Consumption (Some Comparisons Worldqffairs, Summer 1989. p 10. The Affluent Japanese Household, Business America. kh 23,1981, p. 10.
S. Department of Housin and Urban n Survey for the Unrtedktes in 198Q tsrintina Office Julv 18811 The housing conditions of the inner-city generally are similar to the poor throughout the rest of the country. Among the poor living i partments, 4.5 percent live in build ings with some broke n or boarded-up windows. Eleven percent of poor households in 1989 reported signs of rodents within the last three months. Based on Census Y 22 Data on broken and boarded-up windows in the 1989 survey are limited to multi-unit dwellings 14 Bureau criteria, 1 percent of poor central city households had severe upkeep problems, and 6.5 had moderate upkeep problems. But the overwhelming majority of housing units occupied by the poor in the central city are in reasonable condition and are free of such defects TH E SPECIAL CASE OF THE HOMELESS While the overwhelming majority of poor persons identified by the U.S. Census Bureauare well fed and decently housed, the Census Bureau generally excludes h me less persons from its housing surveys and its annual poverty and income surveys.
Nevertheless, this does not alter the data in a statistically significant fashion, even though the homeless problem has clear implications for public policy.
While the Census does not capture the homeless in its annual population and incom e surveys, the highest scientific estimate, compiled in 1989 by Martha Burt and Barbara Cohen of the Urban Institute, places the number of homeless at between 355,000 and 445,0
00. Thus even if the homeless were correctly included in the Census poverty count, they would represent no more than 1.5 percent of all poor persons.
Rejecting the Urban Institutes estimate of the number of homeless persons, and op ting for a higher number, still would not significantly alter the overall picture. Even the politiciz ed and thoroughly discredited claim of two million homeless persons, made by some homeless advocates, means the homeless still would represent no more than 6 percent of all poor persons. Including the homeless in the Census poverty count thus would not al t er the fact that the overwhelming majority of Americas 30 plus million poor persons a~ housed in reasonably good conditions 23 UNSCRAMBLING THE POVERTY PUZZLE The Census Bureau reports on poverty declare year after year that there are over 30 million poor Americans. But other government surveys measuring actual food con sumption, nutritional status, housing, and property ownership show that few of the households defined as poor by the Census are poor in the sense understood by most Americans. Furthermore, in contrast to the official Census poverty count, measures of the actual living standards of the poor show improvement over time. How can the con tradiction between the official poverty numbers and these other data be explained?
There are five reasons for this disagreement 23 The Census Bureau, however, did attempt to count the homeless in the 1990 decennial census of the entire U.S population. In that year, the Census counted 230,000 homeless persons in shelters and on the streets and in abandoned buildin g s 24 Martha R. Burt Developing the Estimate of 500,000 to 600,OOO Homeless People in the United States in 1987 Department of Commerce, U.S. Bureau of the Census, Enumerating Homeless Persons: Methods and Data Needs Washington, D.C U.S. Department of Comme r ce: March 1991 pp.130-139 15 Reason 1 A large number of the poor, as defined by the Census Bureau, are only temporarily poor. This category would include many individuals thrown out of work. Included also are the families of businessmen who suffer losses i n a given year. Such families may have reasonable amounts of savings which they spend during the period of lost or reduced income. Despite temporarily low income, their life-style may not change dramatically. Few Americans would consider these households as poor.y They are radically different from families trapped in welfare dependence and permanent poverty.
Without in any sense minimizing the emotional and economic distress felt by a family whose breadwinner has lost a job, lumping the temporari ly poor a nd long-term poor together into a single category of the poor only creates harmful confusion among policy makers and leads to ill-con sidered policies. Welfare is not an answer for the temporarily poor.
These families would benefit much more from policies that foster a robust economy and new job creation cash earnings which they do not report to the Census Bureau or the Internal Revenue Service. There are large numbers of individuals work ing primarily for cash payments rather than salaries. These include waitres ses, housekeepers, handymen, repairmen, and many other self-employed persons. Census Bureau officials admit that those who have not fully reported cash earnings to the IRS are unlikely to report those same earn ings to Census takers. A recent stud y of welfare mothers in an uniden tified major midwestern city, for example, found that over 90 perc nt had jobs and income they had concealed from government authorities Some middle and upper income families do the same, of course. Govern ment surveys of expenditures show that low income households typical1 spend $1.94 for every $1 .OO of income reported to the Census Bureau.
This explains in part why some families may be reasonably well fed and housed while having little or no apparent income.
Reason The Census Bureau does not count much welfare assistance in its calculation of poverty. A family is defined as poor by comparing family income to an official income level, or poverty threshold. The poverty threshold for a family of four was $13,359 in 19
90. But in exclud ing nearly all welfare assistance from its calculation of family income, the Census Bureau gives the public an erroneous picture of poverty in America.
Total annual federal, state, and local welfare spending amounted to 184 billion in 19
88. This figure excludes programs for the middle class such as Social Security and Medicare. But of this total welfare spending Reason 2: Significant numbers of poor and low-income families have 85 22 25 Christopher Jench and Kathryn Win, The Real Welfare Problem The American Prospect Spring 1990, pp. 31-50 26 See Robert Rector, How Poor Are Americas Poor? op. cit p. 5 16 the Census Bureau counts only $27 billion as income.Thus 158 billion of welfare assistance is devoted to helping the disadvantaged but i s not counted by the Census Bureau. This is equal to an average of $1 1,200 for each poor household.
If the total value of all welfare benefits were counted, many poor families would be shown to have financial resources well in excess of the governments po verty threshold. This, in part, explains why so many wel fare families may be reasonably well fed and housed despite the fact that the Census Bureau reports them to have little or no income Reason 4: The Census Bureau ignores all assets and savings in det e r mining whether or not a family is in poverty. But a family which has fully paid for and furnished a house, and paid for a car, will not necessari ly have a poverty life-style even if its annual cash income is regularly below the official poverty level. T his point is particularly important Reason 5: Low family income does not automatically indicate destitute among the elderly living conditions. There is a general assumption that all households with incomes below the federal poverty income threshold are un a ble to afford adequate food and housing. This assumption is incorrect. Given the low cost of basic food and housing, for instance, it is possible for many families to feed and shelter themselves adequately while having incomes below the poverty threshold.
And many do. The median annual housing costs for the average poor household of two persons, including utilities and taxes, came to $2916 in 1989.2 According to the U.S. Department of Agriculture, a thrifty but nutritious diet for that same household would cost about $1980 per year So, the combined food and housing costs would amount to $4896, or 61 percent of the official poverty threshold in that year for a household of that size. Similarly, food and median housing costs for a poor married couple wi two c hildren living in non-subsidized rental housing is $8267 per year This figure is 65 percent of the poverty threshold for a family of that size. Over half of all poor households, of course, actually have housing costs below the median figures given above. G iven the typical cost of basic food and housing in the U.S it is simply incorrect to assume that most families with incomes below the poverty level must be inadequately housed and fed 2P 21 The American Housing Survey for the United States in 1989, p. 60 2 8 Figures on median housing cost of non-subsidized poor renter households are from the Joint Center for Housing Studies, Harvard University, The Slate of the Nations Housing 1990, op. cit p 21. Food costs are based on the Agriculture Departments mIFN Food Plan. All figures are for 1989 17 CREATING AN ACCURATE MEASUREMENT OF POVERTY It widely is agreed that federal welfare policy clearly is in need of an overhaul.
While the 1988 Welfare Reform Act was designed to change welfare policy substan tially, the la w largely expanded current programs. The perverse incentives that promote dependency, illegitimacy, broken families and hopelessness among the poor remain intact in the system.
Fashioning an intelligent policy to attack poverty is impossible without a cle ar idea of the true dimensions of the problem This means separating fact from fiction, and replacing popular prejudices with hard data. Without that, Congress cannot prescribe an effective remedy to the problems of the poor or target assistance to those l o w-in come families that need the most help. The first step to developing sound reform policies, then, is better information curate information about the earnings and welfare benefits received by these households, its report should be replaced with a new s ystem of assessing poverty in the U.S. This system would have two components 29 Because the Census Bureau's annual poverty estimate does not provide such ac The system should improve existing measures of actual living conditions.
The federal government cur rently uses surveys that measure the actual housing con ditions, property ownership, food consumption, and nutritional status of the poor These reports include Food Consumption Surveys published by the U.S. Department of Agriculture, Housing Characteristi c s, published by the Department of Energy, the American Housing Survey, published jointly by the Census Bureau and HUD, and various reports of the National Health and Nutrition Examination Survey, published by the Department of Health and Human Services.30 These reports contain accurate and valuable data on the living conditions of poor Americans.They paint a very different picture of low income Americans than that con tained in the Census Bureau's annual poverty report. And in contrast to the Census povert y report, they show genuine improvements in the living standards of the poor over time 29 For a general discussion of welfare reform see Robert Rector and Michael McLaughlin A Conservative's Guide to State Level Welfare Reform" in A Conservative Agenda for - the States, published by theTexas Public Policy Foundation, forthcoming, pp. 5-6 30 Examples of these are: U.S. Department of Agriculture, Food Consumption: Households in the United States, Spring 1977, Report No. H-1 Washington, D.C U.S. Government Print ing Office, September 1982 U.S. Department of Energy, Housing Cfwacteristics 1987 Washington, D.C U.S. Government Printing Office, May 26,1989).
National Center for Health Statistics, U.S. Department of Health and Human Services, Anthropometric Reference D ata and Prevalence of Overweight, United States 197880, Vital Health and Statistics, Series 11, No. 238 Washington, D.C U.S. Government Printing Office: 1987 18 CONCLl The main problem is that while the government collects this solid data, few policy make rs apparently read them; some, doubtless, have never even heard of these govern ment surveys.
The federal government thus should conduct these surveys more frequently, expand them to collect more detailed information about low income households, and give t hem much greater prominence. At least once every the years the government also should issue a single combined report on the actual living conditions of the poor, in tegrating hard data from these sophisticated federal surveys The system should measure cor r ectly the financial resources available to The Census Bureaus current method of measuring the income of poor households is low income households grossly inaccurate. The Consumer Expenditure Survey published by the Bureau of Labor Statistics, a bureau of t he Department of Labor, shows that low income households spend $1.94 for every $1.00 of income reported by the Census Bureau.
This survey of actual expenditures gives a far more accurate picture of living standards than the emneous Census income measure. I n addition, poor households receive many non-cash welfare benefits such as medical assistance and housing subsidies that are not measured in the consumer expenditure surveys The government should mate a new survey to measure the actual financial status of poor households based on an accurate count of expenditures and non-cash welfare benefits. This survey could be used to determine the number of households in which expenditures and welfare benefits fell below the poverty income th~holds LJSION The Census B u reau figures on the rate of poverty in the U.S. form the basis of wel fare policy making. But the figures are wrong. The Census Bureau greatly undercounts the incomes of poor Americans, and the actual physical living standards of the households identified poor by the Census Bureau are far better than the public im agines.
Of course, Americas poor do not live lavishly. But few households are poor in the sense of being destitute. The average poor family is well fed and well housed in a two bedroom house or a partment that is structurally sound, in good repair, and well heated. Very few poor persons live in crowded conditions. Forty percent of all poor households own their own homes. And the average home owned by a poor household is a the-bedroom house in good repair with a value equal to 58 percent of the average value of all homes owned in America 31 Robert Rector, Kate Walsh OBeime, and Michael McLaughlin, How Poor Are Americas Poor? Heritage 32 The proposed survey would combine elements of the Consumer Expe n diture Survey and the Survey of Income and Foundation Backgrounder No. 79 1 Program Participation 19 Dramatic Improvement. Moreover, the housing conditions of the poor have im proved dramatically over the last thirty years. In 1960 some 25 to 30 percent o f poor households actually lacked an indoor toilet. Today such situations are virtually non-ex istent. Today 53 percent of poor households have air-conditioning, compared to 36 per ent among all U.S. households, both poor and non-poor, just twenty years ag o.
For state legislators and Members of Congress to devise sound welfare policies, they need accurate information. They do not have it today. The federal government should revise its surveys to provide more accurate information about the financial resource s and actual living conditions of poor Americans. Such good information is the neces sary first step in reforming Americas failed welfare system.
Robert Rector Policy Analyst 20