ECO OF POVERTY AND WELFARE PROGRAMS
ECO OF POVERTY AND WELFARE PROGRAMS ECO 450G
Popular in Course
Popular in Economcs
This 29 page Class Notes was uploaded by Miss Annamarie Kovacek on Friday October 23, 2015. The Class Notes belongs to ECO 450G at University of Kentucky taught by Staff in Fall. Since its upload, it has received 38 views. For similar materials see /class/228266/eco-450g-university-of-kentucky in Economcs at University of Kentucky.
Reviews for ECO OF POVERTY AND WELFARE PROGRAMS
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 10/23/15
ECO 450G Notes Spring 2003 Semester 7 March 13 2003 Chapter 10 Discrimination in the Labor Market I Racial Discrimination in the Labor Market 11 Class Discrimination in the Labor Market 111 Sex Discrimination in the Labor Market I Racial Discrimination in the Labor Market If minority workers are discriminated against in the labor market their incomes will be depressed If discrimination is prevalent at the bottom of the labor market then minorities will be heavily represented among the poor There are substantial gaps in the poverty rates across race gaps that could be explained by a host of factors including discrimination Note that Schiller claims that the whitenonwhite gap in average incomes is 24000 while the number given in chapter 7 was only 17000 Discrimination may manifest itself in a number of ways Nonmarket discrimination Part of the differences in incomes results from differences in schooling and part comes from regional and urban differences Past labor market discrimination Prior discrimination would result in minorities being less skill less experiences and lower seniority Current labor market discrimination As we saw in the Bertrand and Mullainathan study there is credible evidence of discrimination in the labor market today Figure 101 shows that among workers the mean earnings of white males was about 50 higher than for black or Hispanic males 7 about 38000 versus 25000 per year Severa1 questions come up about these number from the 1998 CPS supplement Chapter 7 which was alluded to on page 1745 of this chapter cites incomes of 43000 for white families and 26000 for black hispanic families see page 134 of Schiller first full paragraph The numbers in chapter 7 on the surface seem inconsistent with these numbers because the gap between earnings and family income is only on the order of 10005000 This would seem to imply that either a very few of these men are married b they are married but have nonworking wives The other explanation which is probably correct is that these numbers are only for those who worked at all 7 the unemployed who have zero earnings 7 are probably not included Ifthey were included these numbers would be somewhat lower In Table 102 Schiller cites unemployment rates for whites around 4 Hispanics around 6 and Blacks around 8 Note that the title to figure 101 says the earnings are mean earnings while the last sentence on page 176 first sentence on page 177 says they are median earnings Part of the 50 earnings gap is surely because of differences in education 7 as Table 101 illustrates once education is accounted for the average earnings are more similar in general but are still not equal Again several questions arise with Table 101 7 this table apparently is used for the graphs in Figure 82 on page 146 These numbers then provide further evidence of the typos by comparing Table 81 and Figure 82 Schiller points out that although controlling for years of schooling does reduce the gap in earnings even with the same education grouping there are differences in the quality of schooling Therefore the earnings gap would shrink further if we were able to control for differences in quality The conclusion from these graphs therefore is that much of the earnings disparity between whites and blacks can be attributed to prior nonmarket discrimination Only about onefourth of the disparities are directly attributable to discriminatory labor market practices The conclusion therefore is that most of the problems occur outside of the labor market not inside it To the extent that labor market discrimination exists however one can break down earnings differentials by employment levels occupational differences wage differences and training differences Although each of these gaps helps explain the whitenonwhite earnings differential they neither prove nor disprove discrimination Unemployment rates are higher for blacks and Hispanics Table 102 The duration of unemployment is higher for blacks but not Hispanics There are differences in occupational status e g the distribution of whitecollar blue collar servicejobs and farm workers See Table 103 It is clear from this table however that the differences in occupations by gender is E more dramatic than the differences by race Men are relatively more concentrated in bluecollar jobs while women are relatively more concentrated in the whitecollar and service jobs Holding constant occupation and hours of work Table 104 shows that white men almost uniformly make more money per week than black or Hispanic men To the extent that the jobs with these occupations really are similar this would provide the most conclusive evidence on wage disparities due to race although it still does not control for education differences or job tenure Training Schiller cites a study that whites get nearly double onthejob training as minority workers but does not say what percentage this is There are several ways that discrimination in the labor market may manifest itself The term institutionalized discrimination may be unintended efforts that disproportionately benefit white workers For example wordofmouth recruiting is an efficient way of finding job applicants but may exclude those who are already shut out from the firm Companies may also be less likely to recruit in minority residential areas especially ghetto areas At least at most universities and probably large firms there are very explicit efforts to spread the message about job openings Many universities for example require job announcements to be made in public places for example the intemet and have an application period There is also potentially overt discrimination based on outright racism In the past some labor unions have been guilty of such practices It is also claimed by Schiller page 185 that employers will often ignore potential minority workers not as a result of their own prejudice but because they fear that such hiring will trigger the prejudices of white employees or existing customers This seems inconsistent with many recent debates over affirmative action Many large private companies have come out very publically supporting affirmative action efforts For example in an extremely important upcoming Supreme Court case Gratz v Bollinger and Grutter v Bollinger several white applicants are suing the University of Michigan for using race as a factor in admissions Microsoft Intel American Airlines Procter amp Gamble Eastman Kodak and PepsiCo are among more than 40 Fortune 500 companies siding with the University of Michigan in the case 7 see httpwww 39 quot 39 39 7002070284263001 htm Firms may also have objective but discriminatory hiring practices like the use of employment tests and other credentials These are discriminatory when the tests have no direct relationship to job performance Another practice besides overt discrimination is statistical discrimination Like overt discrimination statistical discrimination is illegal The notion is that because on average white workers have more education skill and experience than nonwhite workers an employer is statistically safer in hiring white workers That is race is correlated with objective characteristics that the employer does want to screen on 7 at least on average It is argued that competitive pressures would leave discriminating firms at an economic disadvantage That is if minority workers were paid wages below their productivity some opportunistic firm could enter the market and make aboveaverage profits Schiller notes that this would be true in a perfectly competitive market but many markets are imperfect 7 there are barriers to entry information imperfections brand name loyalty etc Nonetheless even if competitive practices do not entirely wipe out discrimination they probably put important limits on it H Class Discrimination in the Labor Market The notion here is that equally qualified individuals from low income backgrounds are denied access to betterjobs and pay Corporations may fail to recruit in poor areas and ghettos and may rely on recruiting mechanisms like word of mouth contact Moreover Schiller cites that Conduct speech and dress are among those factors that create communications barriers Employers tend to see these differences as indicators of ability rather than the result of socioeconomic environment It is important to note that the line between what is discrimination is very difficult here Issues like conduct speech and dress are potentially indicators of ability to do a job 111 Sex Discrimination in the Labor Market Women tend to enter the labor market with less human capital training for different kinds of work and potentially different preferences for full versus parttime work As Table 106 shows the occupational distributions for men and women are somewhat different and the presumption is that women are pushed into traditional jobs for women Note that for most jobs there is no biological basis for stating that they should be malejobs or femalejobs The table implicitly defines traditional jobs not by any biological basis but by the sex composition It is therefore true in a mechanical sense that traditional female jobs will have a majority of female employees and traditional male jobs will have a majority of male employees Occupational segregation job interruptions due to childbirth and discrimination are the likely causes of these wage gaps ECO 450G Notes Spring 2003 Semester 7 March 11 2003 Chapter 9 Discrimination in Education I Discrimination ll Racial Discrimination in Education 111 Class Discrimination in Education IV Sex Discrimination in Education I Discrimination Attitudes versus Action Several terms are important Discrimination Unfavorable actions that people harbor against others especially population groups Prejudice Unfavorable feelings and attitudes that people harbor against others especially population groups Racism Encompasses both discrimination and prejudice Discrimination need not convey notions of injustice or injury 7 in general discrimination refers only to the differential treatment of persons As Schiller mentions we discriminate for example when we assign the tallest boys to the basketball team The Relevancy Standard To come up with a relevant concept of discrimination need to recognize on which differential treatment is based Assigning taller boys to a basketball team for example serves a very speci c and productive function Cannot make this kind of argument when assigning jobs or education across race Discrimination based on irrelevant or nonproductive criteria must be considered as injurious to the public welfare US Supreme Court Case 7 in 1971 Court ruled that intelligence tests and other hiring criteria that were not related to job performance were illegal These tests which generally ruled out many minority applicants were a form of discrimination Figure 91 shows that Blacks and Whites have different perceptions about equal opportunity Whites are more likely to answer that yes to the question Do blacks in your community have as good a chance as whites The poll was from 1997 Although there are striking differences in the responses across race e g Whites were more optimistic than Blacks about equality by significant margins a majority of Blacks thought that they had as good of chance as Whites in housing and education but not in employment It would be interesting to see how perceptions change over time Costs and Bene ts of Discrimination Although society as a whole does not bene t from discrimination based on race sex etc it is possible that some groups do bene t When blacks and hispanics are discriminated against in they labor market receiving lower wages than equally quali ed whites Hence white workers who are immunized against competition from minority workers receive higher wages Employers who actually hire minority workers bene t by getting higherquality labor than they are in fact paying for Some whites can lose too Those whites who are in the same occupations at discriminated workers have their wages pushed down by oversupply Figure 92 shows the supply and demand analysis in two labor markets one of which discriminates This analysis assumes that there is a barrier to entry so that a rm cannot enter and hire the lowwage highproductivity discriminated workers Proving Discrimination Economists have given a great deal of attention to guring out how to prove discrimination 7 they key problem is separating whether a rm is treating someone differently because of race or gender or whether the rm is using a productive criteria like educational attainment that may be correlated with race or gender There are several kinds of processes for guring out whether there is discrimination Smoking guns There are cases that involve blatant discrepancies in treatment 7 for example if a minority group was required to take special tests or possess superior quali cations These are rare examples in real life however 391 nference based on end results Schiller argues that if we see highly skewed results like signi cantly fewer minorities in positions of power or graduating from college we can infer that there is discrimination somewhere along the way Many economists would feel uncomfortable with taking for example black white differences in occupations wages or education and attributing that difference to discrimination without further thought There are likely other factors family structure for example that may have nothing to do with schools or jobs that help explain at least part of the difference The key point then is not that differences in wages or education are fully explained by factors other than discrimination but the default that these differences come from discrimination is too strong of a statement to make In general economists like to run experiments where other factors are held constant and the only factor that varies is race If holding all else constant we saw that race affected the likelihood of college admittance or job offers we would have more conclusive evidence of discrimination Experiments and audit studies There is a recent fascinating study done by Marianne Bertand and Sendhil Mullainathan called Are Emily and Brandan More Employable than Lakisha and Jamal A Field Experiment on Labor Market Discrimination See httpgsbuchicagoedupdfbertrandpdf for the actual article In this paper the authors test whether employers discriminate against black job applicants through an unusual experiment They selected 1300 helpwanted ads from newspapers in Boston and Chicago and submitted multiple resumes from phantom job seekers The job openings involved administrative sales clerical and managerial positions and they submitted resumes patterned after real resumes of people who were actually seeking similar jobs The researchers randomly assigned the first names on the resumes choosing from one set that is particularly common among blacks and from another that is common among whites For example Kristen and Tamika and Brad and Tyrone applied for jobs from the same pool of want ads and had equivalent resumes Nine names were selected to represent each category black women white women black men and white men Last names common to the racial group were also assigned Four resumes were typically submitted for each job opening drawn from a reservoir of 160 Nearly 5000 applications were submitted from mid2001 to mid2002 The authors kept track of which candidates were invited for job interviews No single employer was sent two identical resumes and the names on the resumes were randomly assigned so applicants with black and whitesounding names applied for the same set of jobs with the same set of resumes Apart from their names applicants had the same experience education and skills so employers had no reason to distinguish among them Here are the key results Race discrimination Applicants with whitesounding names were 50 percent more likely to be called for interviews than were those with black sounding names Interviews were requested for 101 percent of applicants with whitesounding names and only 67 percent of those with black sounding names There were differences among the nine names associated with black women but not among the names within each of the other groups At the low end the interviewrequest rate was 22 percent for Aisha 38 percent for Keisha and 54 percent for Tamika compared with 91 percent for Kenya and Latonya and 105 percent for Ebony The names chosen for black women were not uncommon they represent 71 percent of all names listed on Massachusetts birth certi cates for black girls from 1974 to 1979 No gender discrimination Within racial groups applications with men s or women39s names were equally likely to result in calls for interviews providing little evidence of discrimination based on sex in these entry level jobs Location The 50 percent advantage in interview requests for white sounding names held in both Boston and Chicago and for both men and women Credentials The likelihood of being called for an interview rises sharply with an applicant s credentials 7 like experience and honors 7 for those with whitesounding names but much less for those with black sounding names Audit Studies Other researchers sent a small number of matched black and white quotauditorsquot to apply for jobs in person Typically though not always the black job seekers were less likely to be invited for an interview or offered a job Those ndings however were criticized because the applicants knew the intention of the study and might have behaved differently In addition the auditors might not have been well matched with the jobs in question they could have been overqualified or underqualified H Racial Discrimination in Education There were two landmark court decisions in education Plessy v Ferguson in 1896 where the US Supreme Court effectively condoned discrimination with a Separate but equal doctrine and Brown v Board of Education in 1954 determined that segregated facilities were inherently unequal Disparate Outcomes As Schiller notes Blacks and Whites go into the education system comparatively equal but come out of the system very different Test scores Nineyearold whites for example score 58 better on reading science and math tests These gaps grow over time Dropout rates Over 92 of whites finish high school only 86 of blacks and 71 of Hispanics Illiteracy A 1975 study showed dramatic gaps in functional illiteracy by race College graduation Graduation rates for whites is about 1015 percentage points higher than for Blacks or Hispanics Schiller labels these facts as indirect evidence on discrimination Do these different outcomes however prove in any way that discrimination is present School Segregation A large percentage of black and Hispanic children attend schools that are highly segregated 7 that is dominated by one race Table 91 shows that most black students attend schools with a majority 50 of minority students though the percentage has fallen over time This is particularly true for blacks attending schools with an overwhelming minority enrollment 7 it has fallen from about twothirds in the late 1960s to about onethird from 1980 onward In part this is because of desegregation efforts in the 1970s 7 the busing of minority students to white schools Segregation levels for Hispanics have not increased rather than declining however In a recent study Jonathan Guryan studied whether the desegregation benefited black students He finds that desegregation plans of the 1970s reduced the high school dropout rates of blacks by 13 percentage points and can account for about half of the decline in dropout rates of blacks between 1970 and 1980 A similar analysis suggests that desegregation plans had no effect on the dropout rates of whites If you are interested you can view this paper at httpwww gsbuchicagoedufac j onathan guryanresearch GuryanDese gpdf Classroom Segregation Even if schools were more integrated classrooms might not be In part integration was achieved by busing black pupils to previously white schools In some cities like Milwaukee the black pupils were maintained in different classrooms More generally it is thought that classrooms are more segregated than schools One form of separating students that Schiller asserts is a subtle form of classroom segregation is tracking 7 where more able students are separated from others and taught differently The reason he claims this is segregation is that IQ tests and other achievement tests are used for tracking Equality of Facilities The Coleman Report 1966 was cited by school districts around the country as evidence that integrating black kids into white schools would have little or no effect on student achievement It explained differences in academic achievement between whites and blacks as a byproduct of a culture of poverty This culture of poverty supposedly had a greater in uence on blacks because of a higher concentration of poverty among blacks Most measured individual differences between black schools and white schools were relatively small Teachers training teachers salaries and curriculum were relatively equal Little difference between predominantly blacldwhite schools in funding buildings age library facilities number of textbooks teacher characteristics and class size The KainSingleton study tracked 18 million elementary school pupils in 4500 Texas schools for five years in the 1990s It did find several measures of school quality that contributed to disparities across race unlike the Coleman Report Surprisingly economists and others have found weak or nonexistent relationships between educational inputs eg smaller class sizes and educational outputs e g test scores and some labor market outcomes One leading economist Eric Hanushek in quotThe Economics of Schooling Production and Efficiency in Public Schoolsquot Journal of Economic Literature September 1986 argues the effects of educational inputs such as per pupil spending teacher experience and teacher degree level have been shown to be relatively unimportant predictors of outcomes and the impact of any particular input to be inconsistent across studies 111 Class Discrimination in Education Poor individuals are more often provided substandard educational facilities School Finances One example is with school finances about half of elementary and secondary school expenditures are financed by local property taxes 7 this creates great disparities both across states and across school districts within a state Many states including Kentucky have passed school finance equalization laws which redistributes funds from rich to poorschool districts Melissa Clark see httpwwwprincetoneduNmaclarldkerapdf studied the Kentucky Education Reform Act KERA implemented in 1990 It included a new funding system to correct large financial disparities between school districts curriculum revision and standardization and increased school and district accountability KERA did successfully equalize perpupil expenditures across rich and poor districts KERA s effects on student achievement have been more mixed Black students in Kentucky have experienced sizeable test score gains since KERA s implementation but the scores of white students have remained unchanged relative to their peers in surrounding states There is no evidence that KERA has narrowed the gap in test scores between rich and poor districts Educational Attainments Poor children drop out of high school at over twice the rate of nonpoor children and some leave before they even enter high school Schiller notes that College admissions are still reserved primarily for those who can support themselves or can afford to forego several years of employment income This neglects to note however that the opportunity cost of attending college is lower for poor students and that many federal grants and loans are meanstested Table 92 shows that college attendance does increase with income but that there are not significant racial differences within income class in attendance In Table 93 Schiller shows a correlation between higher income not defined and college attendance and the same for high ability again not defined Those with low incomes low abilities are the least likely to go on to college and those with high incomes high abilities were most likely to go To see more details on this study go to httpwwwedgovofficesOUSPFS f 391 ll98html or download the full report at httpWWW ed onv offices OUSPFS f ll98pdf Although Schiller does not mention it this Mathtech Inc study reports that of those students who say they do not plan to attend college immediately after high school 57 percent of the bottom income top test score students report that it is because they cannot afford to attend 38 of the middle income top test students give this reason and 21 of the top income top test score students give this reason A substantial number have nancial constraints but the number is far from 100 IV Sex Discrimination in Education The disadvantages that women confront are more subtle 7 steering toward courses that are traditionally female and often lower wage Schiller cites uneven sex ratios in a number of disciplines 7 only 30 of physics chemistry or computer science majors are women It is not always clear however whether these kinds of divergences re ect some barrier to entry e g educational sexism gender bias in testing lack of role models or some re ection of preferences There are no formal requirements for any major related to one s gender Graduate degrees Schiller asserts that female college graduates have had difficulty gaining access to the professional schools that confer the necessary credentials for many desirable jobs Secretarial schools have always been easy to get into but law schools medical schools and business schools have often been a different story There are far fewer women graduating with advanced degrees in some disciplines like engineering that tend to pay pretty well Schiller later asserts that this is a pattern of sex discrimination Page 172 ECO 450G Notes Spring 2003 Semester 7 January 23 2003 Chapter 2 Counting the Poor I The Distribution of Income II The Official Poverty Line III The Number of Poor People IV Measurement Problems V Characteristics of the Poor I The Distribution of Income National statistics collected through annual household surveys One prominent government survey is the CPS Current Population Survey It is not only used by the government to compute the income distribution and unemployment rate but widely used by academic researchers to answer a number of questions about the economy You can see the CPS questionnaire at httpwww hls census 39 htm Using CPS data on approximately 50000 households 150000 people the Schiller book reports that the average family had annual income of roughly 59589 in 1998 median income was somewhat lower at approximately 46737 The numbers reported in the book are somewhat higher than those in official reports see httpwwwcensusgovprod2002pubsp60218pdf which reports numbers that are at least 6 lower As Table 21 of Schiller illustrates the income distribution in the US is very uneven The top quintile top 20 of households receive about 50 of total wealth and the mean income in the top quintile 148846 is more than 10 times that of the lowest quintile II The Official Poverty Line The official poverty line in the US uses the absolute approach to define poverty Thus the government must try to identify the minimum amount of money required to sustain a family In 1998 this was 16660 for a family of four How Was the Poverty Line Obtained Minimum Needs To establish an absolute poverty line we require some notion of what bundle of goods and services is minimally adequate Such a bundle will include food clothing shelter fuel and perhaps other goods In principle we could generate such a list of minimum subsistence based on components like minimum caloric intake etc Units of Measure Once we arrived at some agreement on the subsistence levels for essential items one could try to convert those subsistence levels into money income By doing so we allow for some disagreement in what the actual subsistence levels are 7 individuals could choose differing amounts of the subsistence goods Note that although the income flow is the single best indicator of purchasing power other indicators like access to credit asset levels and inkind assistance are not accounted for by a poverty line that is re ected in annual income CEA Line The first of cially sanctioned poverty line was establed by the Council of Economic Advisers CEA in 1963 at 3000 per year per family The established this line by observing that a minimally adequate diet for a family of four would cost about 1000 per year and that consumer studies indicated that lowincome families spent approximately twothirds of their income on nonfood items Thus the CEA scaled up the 1000 of minimum food expenditure per year to 3000 of minimum total expenditure By their estimates 334 million people lived in poverty in 1963 The CEA measure did not account for family size 7 obviously it is somewhat more expensive to meet minimum needs with larger families and somewhat less expensive with smaller families Thus the CEA measure underestimated the number of large families in poverty and overestimated the number of small families in poverty SSA Index The Social Security Administration SSA revised this measure in a number of important ways these are largely incorporated in the current poverty measure Family size The poverty line was adjusted for family size but this was not as simple division That is POVN POV4 except for N4 Larger families enjoy some economies of scale 7 for example rent and heat typically do not decline by half for a family of two Thus SSA computed equivalency scales 7 the factor that translated a budget for a 4person household into a budget for other household sizes SSA s poverty calculation for a family of four in 1963 was 3130 not much different from the CEA s poverty calculation Current Poverty Index The current poverty line is adjusted annually for in ation To see how in ation has varied over time you can view Table B 60 of the Economic Report of the President available at httpw3accessgpogovusbudgetfy2003pdf20027erppdf The table is also at the end ofthese notes Between 1963 and 1998 price had increase by about a factor of five and the poverty line for a family of four was 16660 The poverty line in the US as it currently stands does not imply an increased standard of living for the poor It only rising price In reality there has been real economic growth over time This means that even after adjusting for in ation the standard of living for most households in 1998 was better than the standard of living for similar households in 1963 The absolute poverty line which is indexed for in ation does not re ect a rising standard of living for the population as a whole Selected Poverty Thresholds in 1998 Family Size 1 8316 4 16660 8 28166 As can be seen from the above table taken from Table 24 of Schiller the poverty line is nonlinear Note that with the exception of Alaska and Hawaii the poverty line in the US is not adjusted for costofliving differences That is it is assume that the same amount of money is needed to meet basic needs in different parts of the country Hard Choices The American public is occasionally asked What is the smallest amount of money a family of four needs to get along in this community In 1996 the median answer was 30000 see Table 13 Schiller therefore concludes that the public s view of minimum needs is well above official poverty thresholds and that most Americans do not believe that the official poverty standard is high enough It is not clear what to make of this survey evidence however In particular it is not clear what goods the survey respondent is envisioning when making this income calculation For example does the respondent view a television microwave oven or automobile as essential Cultural Context In the US the poverty line is an official line that separates poor from non poor It does not necessarily indicate what is enough There is no official de nition or line for the middle class or upper class 111 The Number of Poor People As mentioned previously the Census Bureau uses the CPS to compute poverty statistics annually based on money income See httpwwwcensusgovhhespovertypovdefhtml for more details At the end of this handout a gure from httpwwwcensusgovhhespovertypovert 1pov01chtgif shows trends the number of people and the percentage who live in poverty over time The main things to take away from this chart are that the poverty rate fell dramatically in the 1960s increased in the late 1970s and has remained in the 1215 range thereafter Even during the economic expansion of the last decade the poverty rate peaked at 151 in 1993 and troughed at 113 in 2000 How Poor The poverty gap refers to how far below the official poverty index a family is A family that is slightly below the poverty line is clearly different than a family that is thousands of dollars below that line The aggregate poverty gap is the total income shortfall below the poverty line by all poor people In 1998 it was 83 billion or approximately 2400 per poor person IV Measurement Problems Is the poverty line too low or too high Some believe that it is too low They disagree with the notion that the poverty line should be an absolute measure meaning that it should not increase with the improving standard of living In addition the fraction of the typical budget that is devoted to food expenditure has changed over time Others believe the poverty line is too high For example the poverty line only counts income but not assets such as owning a home There are at least ve difficulties with the current poverty line Inkind income Many poor families receive income I39m kind 7 meaning they receive goods or services directly from the government instead of cash These are clearly valuable For example poor households might receive food stamps Medicaid health insurance Medicare if they are elderly subsidized housing energy assistance and subsidized childcare Because they receive these goods inkind these households do not need to use their own cash to purchase them The official poverty count ignores inkind income Schiller s Table 25 shows that the poverty rate in 1998 would fall from 127 to 95 if inkind transfers and tax credits were included About nine million people would be taken out of poverty Not only is the poverty level affected by inkind income but poverty trends as well Inkind income particularly health insurance subsidies have grown tremendously over the past forty years In 1999 inkind income accounted for 75 of welfare bene ts up 20 percentage points from 1968 At the same time it is difficult to value many inkind transfers The value of health insurance for example depends on health status In general the value of inkind assistance depends on whether the tranfer is marginal or inframarginal To illustrate consider how inkind assistance affects the budget constraint of a household compared to an equivalent amount of cash The figures here are available in more detail at http gattonukyeduF acultyyelowitz479 S PRING2002lecturespdf In the rst budget constraint the individual s decision is unaffected by the cash versus inkind transfer 7 thus he values the inkind transfer at it s cash equivalent In the second budget constraint the decision is affected and the in kind transfer forces the person to overconsume food 7 thus in utility terms the inkind transfer is not as valuable as the cash equivalent Underreporting People may lie to survey takers about their income especially if they believe that their welfare benefits or taxes might be affected or if their incomes come from illegal sources Some estimates suggest that poor families spend twice as much income as they report receiving and the gap between spending and reported income as widened over time Of course access to credit or changing wealth could explain this too Income mobility Many spells of poverty are fairly short There is no obvious reason why income over the course of one year is the appropriate measure Why not one month Or five years Three out of five families that are in poverty in one year are out of poverty the next year Only one in ten families stays in poverty for five years or more Uncounted poor The CPS survey does not interview the homeless those who are institutions and prisoners Schiller argues that some families include unwanted aged relatives who may not share equally in a family s resources and therefore should be counted as poor Schiller calls this disguised poverty It is not clear however why this distinction could not also be made other household structures 7 there are marriages that certainly stay together out of economic necessity for example where one spouse or the other does not enjoy equal access to the resources Latent poverty The poverty measure takes into account cash income from all sources which includes not only earnings interest income dividends etc but also cash transfers such as welfare and social insurance programs like Social Security Obviously many individuals who are collecting Social Security are elderly and have few other sources of income Nearly 40 of all pretransfer poor are kept out of poverty solely because of government checks By pretransfer we mean counting income sources such as earnings pensions etc but not counting government payments By far the largest part of this reduction in the official poor is because of Social Security payments Of 269 million who are kept out of the official poverty rolls by government transfers about twothirds receive Social Security Of course people pay into Social Security during their working lives Ifwe are to exclude Social Security benefits one should presumably consider what would have happened to those Social Security contributions during the working life Presumably at least some of those contributions would have been invested in the absence of Social Security This means that private income sources like pension income and dividends would have been higher too V Characteristics of the Poor Age and Family Status Poverty rates vary tremendously by race blacks and Hispanics have poverty rates around 25 whites have rates around 8 Female headed households have very high poverty rates 7 around 35 The elderly tend to have lower poverty rates than the national average due in large part to the generosity of Social Security Geography The Northeast and Midwest regions tend to have lower poverty rates than the national average the South and West have higher poverty rates Interestingly despite the discussion of the South in the text the poverty rate in 1998 was actually higher in the West Of course these estimates do not account for costofliving differences presumably this would overstate poverty in the South because costofliving tends to be lower Notice that in 1998 however that the poverty rate in the South was actually lower than that in the West Historically if you go to httpwww census vhhespovert 39 vhstpov9html you will see that the South has had the highest poverty rate but the South made relative economic progress in the 1990s Labor force status The de nition of in the labor force is a person who is either employed or actively seeking employment as well as those who are temporarily not working because of illness bad weather vacation or a labormanagement dispute Out of the labor force is everyone else and includes those who are keeping house attending school unable to work because of age or disability or not actively seeking employment Table 29 shows the unsurprising result that for families with children which implies that the head is almost certainly nonelderly virtually all nonpoor families had a working member while a much smaller fraction of poor families had a working member mm Number of Four and Faveny kale 1559 m 2001 32 9 when n 7 Damn W MumWWW ulnmmmumwyun W vmum Nguumuxvrhum mm H WWW yWwwwva wva m w WWWWWW PRICES TABLE B7607C0n5um2rp1i 2 171469625 far majm Expanditmz 145525 19587200 For 311 when consumers 1982784100 except as noted Fund and AM I Devevages H vansr R tE d 01th V93 m mum 19m uusr puvr wear you an gnu 3 CPMU mg tar Hunz Cummumr an we Fund Hun Canon sevvmes 1958 Z89 30 46 Z86 1959 Z91 Z97 450 Z98 1960 Z96 300 457 Z98 1961 Z99 304 46 301 196Z 30Z 306 46 308 1963 306 311 46 309 1964 310 315 47 314 1965 315 3ZZ 47 319 1966 3Z4 338 49 3Z3 1967 334 341 51 308 333 351 1968 348 353 53 3Z0 343 369 1969 367 371 56 340 357 387 1970 388 39Z 59 364 375 409 1971 405 404 61 380 395 4Z9 197Z 418 4Z1 6Z 394 399 447 1973 444 48Z 64 41Z 41Z 464 1974 493 551 69 458 45 8 498 1975 538 598 7Z 507 501 539 1976 569 616 75 53 8 551 570 1977 606 655 786 574 590 604 1978 65Z 7Z0 81 Z4 617 643 1979 7Z6 799 4 701 705 689 1980 8Z4 868 811 831 75Z 1981 909 936 904 93Z 8Z6 198Z 96 974 69 970 911 1983 996 994 995 993 1011 1984 1039 103Z 1036 1037 1079 1985 1076 1056 1077 1064 1145 1986 1096 1090 1109 10Z3 1Z14 1987 1136 1135 114Z 1054 1Z85 1988 1183 118Z 1185 1087 1370 1989 1Z40 1Z51 1Z30 1141 1477 1990 1307 13Z4 1Z85 1Z05 1590 1991 136Z 1363 1336 1Z38 1716 199Z 1403 1379 1375 1Z65 1833 1993 1445 1409 141Z 1304 907 855 19Z9 1994 148Z 1443 1448 1343 9Z7 888 1985 1995 15Z4 1484 1485 1391 945 9ZZ Z069 1996 1569 1533 15Z8 1430 974 953 Z154 1997 1605 1573 1568 1443 996 984 ZZ48 1998 1630 1607 1604 1416 1011 1003 Z377 1999 1666 1641 1639 1444 10Z0 101 Z Z583 Z000 17ZZ 1678 1696 1533 1033 10Z5 Z711 Z001 1771 1731 1764 1543 1049 105Z Z8Z6 Z000 Jan 1688 1661 1660 1483 10Z3 10Z7 Z647 Feb 1698 1663 1671 1497 10Z5 10ZZ Z667 May 171Z 1665 1678 1534 10Z9 10Z0 Z680 AD 1713 1666 1679 15Z9 10Z9 1018 Z719 May 1715 1673 1681 1531 1031 1018 Z70Z June 17Z4 1673 1696 1557 1034 1015 Z696 My 17Z8 1681 1706 1550 1037 10Z0 Z7ZZ Aug 17Z8 1687 1709 153Z 1039 10Z8 Z716 Sept 1737 1689 1714 1547 1038 10Z9 Z747 OS 1740 1691 1717 1544 1038 1036 Z730 NW 1741 1689 1716 155Z 1037 103 Z Z76Z BC 1740 1700 1719 1544 1037 103 6 Z740 Z001Jan 1751 1709 1741 1544 1041 103 9 Z759 Feb 1758 1713 1747 1549 1043 1040 Z77Z May 176Z 1717 1754 1539 1043 1043 Z777 AD 1769 1719 1754 1561 1050 1041 Z813 May 1777 17Z5 1759 159Z 1050 1040 Z80Z June 1780 1730 1773 1583 1048 1044 Z81Z My 1775 1735 1776 1544 1050 1048 Z858 Aug 1775 1739 1780 1533 1051 1058 Z833 Sept 1783 1741 1774 1555 105Z 1066 Z878 Oct 1777 1749 1767 15Z3 1053 1071 Z856 NW 1774 1746 1769 150Z 1055 1070 Z89Z Dec 1767 1747 1769 1485 1053 106 64 mcmdes amuhuhc Devevages not shown sepavate y ZDeCemDev 1 3Huusehu1d Me sigas wed mummy 191 um etcand mutuv m1 Mutuv um CUU aHL etc 3130 mcmded Waugh 1982 NotewataDegmnmg1983mvauvate a mum eqmva enue measme fuv humeuwnevs Busts 391193 ve eut changes My Cumpusmun and venammg Degmmng m 1998 and fuvmma and methudmugy Changes Degmmng m 1999 Some Depavtment of Labuv Buveau of Labuv 31mm Figure 5 14 Example with no inefficiency from in kind transfer Clothes C F Food CF The person consumes more food than the in kind transfer gives him Clothes C Figure 5 15 Example where there is inefficiency Bundle with income transfer Bundle with inkind transfer n F F Food F The person would consume less food with an equivalent income transfer ECO 450G Notes Spring 2003 Semester 7 February 11 2003 Chapter 4 The Working Poor 1 Work Experience and Poverty 11 Searching for Explanations 111 Minimum Wage Jobs IV Low Wages V Why Wages are so Low VI Summary VII Does Prosperity Trickle Down I Work Experience and Poverty There are obviously a number of poor who are subemployed 7 having either parttime or part year jobs which provide insufficient income to lift the family out of poverty Obviously moving out of poverty depends on the intensity of work and the wage opportunities Weeks of Work and Hours In principle one could trace out poverty rates by annual hours of work The CPS asks questions about the number of weeks employed in the last calendar year and the usual hours worked per week when employed Unfortunately annual hours of work in the CPS can only be approximated because usual hours of work per week may refer to the median or mode hours of work or something else rather than the mean For example suppose that a person worked parttime 20 hours for slightly more than onehalf of the year and fulltime for slightly less than onehalf of the year Then usual hours could be 20 hours the median and mode 30 hours the mean or perhaps 40 hours if the person s most recent job was fulltime The questionnaire for the 1998 CPS can be found at J IJ httpwww census V We 98gues98pdf The exact question reads In the weeks that you worked how many hours did you usually work per week See Question 41 of the survey Obviously the word usually can be interpreted in different ways and it is not clear how this question accounts for second jobs temporary jobs Schiller s Figure 41 shows that poverty rates decline rapidly for both parttime and fulltime workers as they work a greater portion of the year Poverty rates are lowest for fulltime full year workers There are a number of issues with this table 1 Poverty rates actually increase for parttime workers going from ll3 weeks of work to 1426 weeks It is hard to know what to make of this increase 7 it may be possible that heads with ll3 weeks of work are different in some way eg family circumstances than those with 1426 weeks This is purely speculation however Notice that at this level fulltime workers also have higher poverty rates than parttime workers Again it is not transparent why this should be so 2 Poverty rates for those with zero weeks of work are not shown 3 The work behavior is for the head of household and does not account for family circumstances Perhaps the most important family circumstance would be between married and single househlds 4 Parttime is de ned as 134 hours per week fulltime as 35plus Since a substantial fraction of full time jobs are centered at 40 hours per week it is fairly clear how to translate fulltime and employed weeks into annual hours On the other hand there is a great deal more dispersion in parttime hours The Working Poor Less than half of the poor household heads 43 have any work experience during the year Of those working heads however twothirds have a full time job 35plus hours per week Schiller claims the presence of so many workers among the poor is a blatant contradiction to the accepted wisdom that to get ahead get a job One should note from his pie chart in Figure 42 however that only a small minority 142 of poor household heads have a fulltime full year job It could certainly be the case that many of those poor households would eXit poverty if they increased their work effort How Much Work There are different classifications of the working poor Department of Labor definition 7 Persons who have devoted 27 weeks or more to working or looking for work and who lived in families with incomes below the official poverty threshold Gorham and Harrison definition 7 All workers whose annualized earnings are too low to lift a family of four out of poverty As Schiller mentions both definitions have problems More Measurement Problems 1 Usual hours per week 7 A person classified as fulltime may occasionally experience part time work Around 10 of fulltime fullyear workers actually had at least siX weeks of part time work 2 Tips commissions bonuses 7 Schiller incorrectly claims that The annual Census survey inquires about regular wages but neglects tips commissions and bonuses These irregular forms of income may be particularly important for workers with low hourly pay eg salespeople and food servers This is incorrect See for yourself in the 1998 CPS httpwww census V awd 39 J 98gues98pdf Question 48a3 and 48aad explicitly discuss this How much did you earn in tips bonuses overtime pay or commissions from this employer in 1997 3 Selfemployed heads 7 nearly 800000 poor working heads out of 5400000 poor working heads classify themselves as selfemployed Computing the wellbeing of the selfemployed is difficult for a number of reasons First there is more potential to underreport income Not all income receipts are easily audited by the Internal Revenue Service for example Second the selfemployed may own assets of considerable value 7 e g a farm or business capital The reason why assets matter is because the selfemployed may have considerable discretion to reclassify salary income into business capital Consider a female head who runs a childcare center out of her home and either a renovates some of the rooms in her home and calls it a business expense or b pays herself a salary and then uses the salary to renovate the rooms In either situation she is fundamentally in the same economic situation but in one case she would have much higher income than in the other Secondary Workers Most individuals live in families and families may supply more than one worker to the labor market Figure 43 illustrates that a secondary worker is often decisive in keeping a family out of poverty In twoparent families with children the poverty rate is 59 when there are no workers 175 with one worker and 28 with two workers A number of issues come up here as well First the classification of workers again does not tell us much about intensity of work 7 presumably the poverty rates are lowest when the workers are fulltime fullyear Second the question arises on why so few families with zero workers presumably over the course of the entire year are not in poverty Schiller cites a study by Levitan that computed that 50 to 70 percent more two earners families would be in poverty without the earnings of the secondary worker There are several things to note about that 7 the poverty rate would still be extremely low because the base of 28 is very low 7 an increase in 50 would lead to a poverty rate of about 2814 or 42 In addition these calculations almost surely assume no behavioral response 7 e g that if the secondary earner left her job that the primary earner would not increase his labor supply This kind of criticism is the same as the latent poverty criticisms from earlier chapters II Searching for explanations Why are fulltime fullyear workers poor Schiller makes several points Obviously earnings is equal to wages multiplied by hours or wH IfH is high then w must be low He also mentions that it might also be the case that the poor simply have aboveaverage needs due to either larger families or special expenses for example medical bills We have learned that larger families will result in a higher poverty threshold so that part of the statement is correct Other special expenses however do not enter the poverty line The poverty line simply measures income ows against a dollar threshold adjusted for family size and in ation 111 Minimum Wage Jobs The nominal federal minimum wage has increased over time from 025 in 1938 to 515 in 1997 It is not indexed to in ation and only increases when there is political agreement on doing so Note that a number of states have by their own choice imposed minimum wages higher than the federally mandated minimum wage See httpwwwdolgo 39 ica htm for statebystate variation Kentucky s minimum wage is the federal one of 515 The state of Washington currently has the highest minimum wage of 701 per hour Living wages 7 a number of cities localities have imposed minimum wages on some subset of workers oftentimes those employed by a company that is dealing with the local government The proposed wage rates are oftentimes around 10 per hour See httpwww enionline m quot 39 inde cfm for an overview The highest living wage is currently in Santa Monica California with a rate of 1425 per hour for employees without benefits See httpwww enionline on quot 39 monica cfm for more details As Schiller notes a fulltime fullyear job of 2000 annual hours would result in income of 10300 which is slightly below the threshold for a twoperson family but ignores a rather large potential EITC refund Figure 44 shows that the real value accounting for in ation of the minimum wage has generally eroded over time There are a number of issues to take away concerning minimum wages and poverty 1 Many people who earn the minimum wage are not in poor families 7 for example many teenagers and secondary workers Only 20 of minimum wage workers are in poor families 2 Wages often increase rapidly for those who start off at the minimum wage 7 these jobs are stepping stones to higher wage jobs 3 Raising the minimum wage potentially has employment effects if the minimum wage is above the marketclearing wage This is a source of great discussion in the economic literature 7 with different studies coming to dramatically different conclusions on the extent of the disemployment effects Card and Kruger find in an in uencial 1994 article in American Economic Review that raising the minimum wage in New Jersey did not have disemployment effects using Pennsylvania as a control Neumark and Wascher in a comment argue that there are disemployment effects when better data is used to measure hiring IV Low Wages One wage earner in a family of four would need an 8hour job to escape poverty Table 43 in the text shows annual earnings of the head of poor families Note that a surprisingly high number of fulltime fullyear heads 127 report earnings of less than 2000 translating into an hourly wage rate of about 1hour These clearly suggest that there is measurement error in either earnings or annual hours Moreover if we divide earnings by fulltime fullyear hours of work about 411 of poor heads are apparently earning less than the minimum wage This seems implausibly high Table 44 shows the distribution of wage rates by age groups nearly 25 million workers were paid less than 8hour Among low earners many were able to escape poverty however The most common way that low earners escaped poverty is by having a small family size and other common ways included having other workers in the family unearned income and welfare benefits Poor Jobs The reported occupations of poor heads of household are concentrated in a number of areas 7 the most common of which are Other service workers Craftspeople and precision production Salesworkers and Clerical workers V Why are Wages so Low Economics would suggest that wage rates are determined by supply and demand There is ample supply of workers for unskilled jobs hence wages remain low Schiller argues that the demand side of the market is neglected in antipoverty discussions and by doing so ignores tremendous potential for eliminating poverty He says that the distribution of wages and incomes is partly a re ection of collective social decisions regarding the merits of particular kinds of output Had we decided instead to dredge more rivers to build more houses or to clear up our cities the extent and nature of poverty might now be marked differen Again Schiller s analysis assumes no behavioral responses Ifunskilled occupations paid more one could reasonably expect that skilled individuals would shift into them For example child care workers and social workers are often paid very little 7 yet might be rewarding in a number of ways One might expect that more skilled individuals would shift into occupations like those if they were more handsomely rewarded VI Summary Many people are poor despite relatively high attachments to the labor force 7 regardless of the measurement issues VII Does Prosperity Trickle Down Do the poor bene t from general economic growth or do they live in an isolated subeconomy The trickle down perspective Asserts that people at the bottom of the economic hierarchy benefit from increased prosperity as higher income levels because of multiplier effects The dual labor market perspective Asserts that a variety of barriers such as discrimination either overt or in a statistical sense or unions reduce the bene ts from growth at the top of the income distribution