This is the fifth and final installment in a series of studies produced for the Joint Economic Committee on the effects of the welfare state on various aspects of the American economy and society. This study focuses on the impact of the welfare state on the American family, including its effects on family structure, family income, children, and the social institutions supporting family life.
As in the previous studies in this series, the excessive size of the federal government is found to be counterproductive, and associated with a variety of negative economic and social consequences. As the size and scope of government has grown, it has undermined the strength of the family and the social institutions that support it. Though well intended, in many cases this excessive expansion of the federal government has produced results the opposite of those intended.
According to this study, a level of federal government spending above about 18 percent of Gross Domestic Product (GDP) undermines economic and income growth; federal spending above this level has depressed family income growth in the period since 1973. The drag on the economy imposed by additional federal taxes and spending has contributed to slowing the productivity gains on which any sustained growth in family income depend.
Moreover, the evidence suggests that younger families are especially vulnerable to the economic costs imposed by excessive government activity. Families headed by adults between 25 and 34 years old suffer disproportionately heavy income losses from counterproductive federal spending. For example, federal spending increases of $100 billion would reduce these families' incomes by an average of $1,418, compared to $397 for a family with adults in the 55-64 age bracket.
Unfortunately, the level of federal government spending is projected to remain at counterproductive levels for the foreseeable future. However, restraint of federal spending and taxes would stimulate additional economic and income growth. According to this study, $100 billion in federal spending restraint would increase real median family income by $895, or by 2.3 percent. The bottom line is that federal spending restraint would not only improve the federal budget outlook, but would improve the outlook for the family budget as well.
America is now engaged in a national debate over whether more of the resources earned by American families should be used as they see fit, or used in Washington to finance excessive federal taxes and spending. This study provides evidence that the shift in resources towards the federal government has simply gone too far, and that federal restraint in taxes and spending would provide more economic and family income growth over the long run. Ultimately, the choice is between a higher standard of living for families, or for government.
It is my hope that this new JEC study contributes to a lively and informative debate.
What has happened to the economic status of middle America? The answer to that question depends on the time period you are talking about. Measured by real median family income, the last few years have witnessed meaningful decline. See Figure 1. In 1994, this indicator of economic well-being is estimated to be 5.2 percent less than it was in 1989.[1] With the rather low December, 1994, to December, 1995, growth in Gross Domestic Product (only 1.3 percent), the 1995 data would be expected to reinforce this pattern. Significantly, the year 1995 is thought by many as being something of a business cycle peak, similar to 1989. At least, President Clinton's claim that the economy is the healthiest it has been in three decades would suggest this.[2]
Six years of net decline in real median family income has dealt a significant setback to middle America.[3] However, if we consider a somewhat longer period of time, the picture is not so grim. In the seven years preceding 1989, real median family income rose by 12.6 percent. Of course, it could be argued that this growth is something of a statistical artifact since the initial year for this comparison is 1982, at the trough of a serious recession. Actually, the distortion from this is minor. If we substitute 1981, a year containing an official business cycle peak, for 1982, the growth in real median family income only falls to 11.0 percent.
This might indicate that the years since 1989 are something of an aberration. However, if we look back beyond 1981, we see more of what has transpired since 1989. In the two years between the business cycle peaks of 1979 and 1981, real median family income fell by 7.4 percent. And before that, from the business cycle peak year of 1973 through 1979, it rose by only 3.7 percent. Over the entire period 1973-1981, this measure fell by 2.7 percent. Thus, if anything, the period 1981-1989 is the aberration, an interval of rising income sandwiched between two substantial episodes of stagnant or declining family economic well-being.
There is more to this story of the behavior of real median family income. While the income growth that characterized the period 1981-1989 was a welcome relief from the stagnation or declines that surrounded it, an examination of the data for earlier years indicates that it pales in comparison. For example, in the six years from 1967 to 1973, real median family income rose by 16.8 percent, in contrast to the 11.0 percent rise in the eight years 1981-1989. On an annual basis, this is about twice the growth rate compared to the 1981-1989 period.
Perhaps the most dramatic demonstration of the changing nature of the patterns of growth in real median family income is to contrast the 21 years from 1973 through 1994 with the 26 years beginning in 1947 and ending in 1973. Between 1973 and 1994, real median family income rose by a mere 2.5 percent, despite the 1981-1989 growth surge. On the other hand, between 1947 and 1973, real median family income increased by 103.8 percent, slightly more than a doubling. See Figure 2. This is a remarkable difference, indicating a marked slowing in family income growth after 1973.[4]
Not surprisingly, the pattern of real family income growth parallels that of real wages and the average productivity of labor.[5] This implies that our previous explanations of real wage and productivity behavior will be pertinent to the issue of why the rate of growth in real family income has slowed so precipitously since 1973. To confirm this, we have conducted the necessary regression analysis, using real median family income as the dependent variable and Federal Government spending (as a percent of Gross Domestic Product), Federal spending squared, and the average productivity of labor as independent variables. Why do we include the average productivity of labor? There are two reasons. First, to the extent it explains the behavior of real median family income, we can calculate an indirect impact of Federal spending on real median family income from the already known effect it has on labor productivity. Second, it enables us to determine whether Federal spending has impacts above and beyond those associated with the labor productivity measure. For example, if the regression results show a significant relationship between real median family income and Federal spending, after controlling for movements in labor productivity, this would indicate an additional impact.
The regression results are reported in Table 1. They do show a statistically significant effect of Federal spending beyond that produced by movements in the average productivity of labor. Recalling the analysis reported in the second of these reports,[6] we found that Federal Government spending greater than 17.42 percent of GDP exerted a negative effect on the productivity of labor. According to the results reported here, that effect is reinforced whenever Federal Government spending exceeds 17.97 percent of GDP. Thus, the impact of Federal spending on real median family income is something of a two-edged sword. Whenever Federal Government outlays rise above the 17 to 18 percent of GDP, the effect on real median family income is to depress it, both directly and through its impact on the average productivity of labor.
It has been standard practice in this series of reports to ask the question, "What would be the impact of $100 billion of Federal spending restraint on the economic measures being discussed?" In this case, the response to that query must come in two parts. First, we need to calculate the indirect effect produced through the impact of Federal spending on the average productivity of labor. We have already shown that $100 billion of Federal spending restraint would augment the average productivity of labor by 0.8 of one percent.[7] That translates into a $320.85 increase in real median family income, at 1994 levels. See Table 2. This is what we have called the indirect effect on family income.
On the direct side, $100 billion of spending restraint would reduce Federal spending as a percentage of GDP by about 1.5 percentage points, assuming 1994 levels of spending. Making the necessary calculations shows that such a change in Federal spending levels would lead to a $574.50 increment to real median family income. This gives a combined effect of $895.35. Such an increase in real median family income is nearly equal to the $944 total rise in real median family income over the years 1973-1994. It would also amount to a 2.3 percent addition to the 1994 level of real income.
More information is available than that for just real median family income in the aggregate. Specifically, the available data describe this statistic by age of the householder in the family. The age categories of interest are 15-24, 25-34, 35-44, 45-54, and 55-64. This makes it possible to replicate the aggregate analysis with the age specific data. The results of doing this are shown in Table 3. They provide some very interesting insights into the impacts of Federal spending on different age groups in America. First, the statistical results all have one thing in common. Real median family income is significantly related to the economy-wide average productivity of labor, although the relationship differs quantitatively. As the age of the family householder increases, the sensitivity to productivity changes rises through the age group 45-54 and then declines in age category 55-64.
When it comes to the Federal spending variables in the regression results contained in Table 3, the pattern is much more disparate. Through the three age groups embracing the years 15-44, the Federal spending variables are statistically significant with signs consistent with the overall results shown in Table 2. However, in the age groups 45-54 and 55-64, the statistical significance disappears.
How can these differences be summarized? We have done it by conducting the $100 billion in Federal spending restraint analysis for each of the six age groups. Again, the effects are partitioned into an indirect one operating through the productivity variable and a direct additional impact created by variations in Federal spending levels. The detailed results are shown in Table 4 and their combined effect is illustrated in Figure 3. Of course, the indirect effect mirrors the already discussed pattern shown in the regression coefficients for the productivity variable. It starts small and then rises, peaking in the age group 45-54, before falling.
Almost exactly the reverse is shown with the direct effects. It starts large, declines through age group 35-44, is not significant for ages 45-64. In combination, the two effects show a powerful pattern of systematic decline as you move to older and older age groups. This implies that the effects of Federal Government spending beyond what we have called the "optimal," or threshold level, is most strongly felt by families with young householders. If the impacts are expressed as percentages of 1944 levels of real median family income, the disparities are even more pronounced. For those families with a householder aged 15-24, we estimate that the typical percentage increase in real income associated with $100 billion of Federal spending restraint would be 10.2 percent. Among those in the peak earnings years, ages 45-54, the relative increase is only 0.9 of one percent.
The findings of the previous sections have some interesting implications for the notion of the American Dream, the idea of persistent intergenerational economic progress, the hope that one's children will have a better life then their parents. If the incomes of families headed by younger householders are more negatively affected by the expansion of the Federal Government beyond its optimal size, their economic status may be adversely affected on a permanent basis. To illustrate the potential magnitude of these effects, consider a younger family, say with a householder in the 15-24 age category, with very young children in 1947. Perhaps 20 years later, in 1967, those children themselves may be householders in that same 15-24 age group. What do the data say about their relative economic status? Simply this; the real income (measured by real median family income) of the 1967 family was 66.1 percent greater than that for the householder's parents in 1947. The American Dream worked.
Move on to 1974. Making the same assumptions, we can compare the 1974 family with one in 1994. The result? In 1994, real median family income where the householder is aged 15-24 is 27.6 percent less than the incomes of a comparable family in 1974. Rather than being better off than their parents, the 1994 family is decidedly worse off. The American Dream no longer exists. Today's families are starting off at a much lower place on the economic ladder than their parents did. See Figure 4.
Perhaps the age group 15-24 is an unusual case, overly influenced by changing patterns of educational activity, for example. To deal with this possibility, we replicate our analysis, focusing on the 25-34 age group. Between 1947 and 1967, the real median family income of families in this age category rose by 83.7 percent, once more confirming the notion of intergenerational progress. However, this is not the case over the interval 1974-1994. Real median family income falls by 7.5 percent, less than the decline in the case of the age 15-24 group, but nevertheless a decline.
We also have calculated the intertemporal changes in median family income assuming a 25 year generation to generation span for the age 25-34 group. The increase in real income for the families with this characteristic was 107.6 percent between 1947 and 1972, compared to a decrease of 4.3 percent between 1969 and 1994. Once more we find a lack of generational progress in the contemporary data.
Perhaps, though, this is compensated for by more rapid increases in income through time. To explore this possibility, we have constructed a hypothetical income profile for those families with a householder aged 15-24 for the twenty year period 1994-2014. To do this, we used the percentage increases in income for a cohort of families for the most recent ten year period for which we have data. To illustrate, we compare the real median family income in 1984 for families with a householder aged 15-24 with the 1994 real income of families with a householder aged 25-34. Once a percentage increase in real income has been determined, we use it to estimate the real median family income in 2004 of a family in the age 15-24 group in 1994. Similarly, we calculate the percentage increase in income as families moved from the age 25-34 to the age 35-44 classification between 1984 and 1994. That is then used to approximate an income level for our age 15-24 family in 2014. The results are shown in Table 5.
An even more dramatic comparison emerges if the 1994 cohort's hypothetical income history is extended to the year 2034 and is then contrasted with the income history of the 1947 cohort over a forty year period, which would take it through the year 1987. At that time, when the 1947 cohort is approaching retirement, the real median family income for their age group, 55-64, was $43,167. Ten years earlier, in 1977, when the 1947 cohort was in its peak earnings years, their real median family income stood at $48,950. Compare these numbers with our projections of the 1994 cohort's real median family income levels for 2034 (age 55-64) and 2024 (age 45-54). The 2034 estimate is $39,435 and that for 2024 is $44,403. Both are significantly less than the actual incomes of the 1947 cohort some 47 years earlier. These data suggest that approximately two generations of Americans have missed out on the American Dream. It appears that the 1994 cohort stands to have a lower level of real income during their peak earnings years than did their grandparents in the 1947 cohort.
As we mentioned earlier, there may be problems associated with basing things on the 15-24 age group. Therefore, we have done a similar analysis in which the 1994 cohort is defined as consisting of those families in which the householder was aged 25-34 in that year. This takes us through the year 2024. The results are shown in Table 6, along with the actual income levels 20, 30, and 40 years earlier. These data tell a somewhat more muted story. The projected income history for the 1994 cohort very closely approximates that for the 1974 group. However, compared to earlier cohorts, those for 1964 and 1954, some income gains are shown. This indicates that perhaps only a single generation, the one currently in the early years of its income life cycle, will be denied the generational gains associated with the concept of the American Dream.
Whether it be one or two generations who have been affected in this fashion, the question remains, "Why?" The earlier analysis suggests that a major explanatory factor is an oversized Federal Government. To confirm the impact of Federal Government spending in this regard, we have attempted to explain statistically the rate of growth in real median family income as various cohorts move through their income history. Figure 5 shows the history through time (1961 to the present) of the ten-year percentage change in the real median family income as people age. Five-year moving averages have been used in order to smooth the data series and minimize the impact of single-year variations.[8]
Data are provided for four different categories, movements between age groups 15-24 and 25-34; 25-34 and 35-44; 35-44 and 45-54; and 45-54 and 55-64.[9] When these data are used as dependent variables in regression equations that have the usual set of Federal spending measures, spending as a percentage of GDP and the square of spending measured in this way, the results shown in Table 7 are obtained. They show what, by now, is a familiar pattern. Up to a certain level of Federal spending, there is a positive impact on the rates of growth in real median family income as householders age and move through their life cycle of income generation. This holds true for the age groups through age 54. In the case of the percentage growth in income as householders move from the 45-54 to the 55-64 age bracket, the statistical results do not show a significant relationship between Federal spending and the dependent variables. In the other three cases, we ask the usual question, "At what level of spending does the negative effect emerge?" The answer is a familiar one, at approximately 17.5 percent of GDP. Once again, we find that Federal Government outlays in excess of about 17.5 percent of GDP are counterproductive.
The quite variable impacts of the of Federal spending on the income history of families has implications for the degree of inequality in the distribution of income in the United States. For example, the fact that the negative effects of excessive Federal spending are more apparent among the younger age groups implies a widening spread between their incomes and those in their peak earnings years. In 1994, the mean (not median) family income was much lower in the age groups 15 through 44 than in the 45 and older cases.[10] While on the subject of income inequality, it is worth noting that, until now, little has been said along these lines. Our almost exclusive emphasis has been on economic efficiency, not equity considerations.
The lack of treatment of equity, or "fairness" considerations could be offered as a criticism of our work. Much of the standard justifications for Federal Government programs and expenditures focus on matters of this sort, arguing that an important function of government is to intervene in the economic system to promote greater equality of economic outcomes. In short, much of the rhetoric in support of government spending is redistributionist in tone. Often, these arguments suggest that there is a conflict between the economic goal of efficiency and a broader social goal of fairness. We will explore that possibility at this point.
The first question is one of measuring the degree of fairness in an economy. While there are some technical ways to do this, such as using Gini coefficients, ordinarily the language of this issue focuses on less technical and more straightforward notions. References to the "rich" and the "poor," the "privileged" and the "underprivileged," the "top" and the "bottom," and "haves" and "have-nots," abound. While these concepts are usually employed in an imprecise fashion, we can devise a measure in the spirit of such notions by simply dividing the income received by a group of people at the top of the income distribution, say those in the top five percent, by the income obtained by those at the bottom of the distribution, in this case, perhaps, the twenty percent of people with the lowest income.[11]
The behavior of this statistic in post-World War II America is shown in Figure 6. It begins at about 3.5 in the late 1940s, falls to between 2.5 and 3.0 in the late 1960s and early 1970s, and then begins a fairly steady ascent until it is in excess of 4.5 by 1994. The 1994 value is the highest in this interval. The question for us is whether the behavior of this statistic bears any relationship to Federal Government spending. To pursue this, we have estimated our standard government expenditure model, with the income distribution ratio as the dependent variable and the Federal Government share of GDP and its square as the independent variables. The regression results are shown in Table 8.
What they show is truly remarkable There is a significant statistical relationship between the income distribution measure and Federal spending. In itself, that is not surprising. It is the nature of the linkage that is striking. At relatively low levels of Federal spending, a rising Federal share of GDP reduces the relative gap between the share of income received by the top five percent of the income distribution and the bottom twenty percent. However, beyond some critical level of Federal Government spending, additional outlays widen the gap. The threshold level where this happens? Where else but in the vicinity of 17.5 percent of GDP. Specifically, it occurs at 17.43 percent, almost exactly the same as the 17.57 percent threshold value for maximizing GDP and the 17.42 percent estimate for maximizing the average productivity of labor and real compensation of workers.
We call this finding "startling." Why? The answer is simple. Apparently, when it comes to the optimal size of government, there is no inherent conflict between economic efficiency and equity, as we have measured it. A Federal Government that is larger than is optimal in an economic efficiency sense also produces a greater gap between the incomes at the top and the bottom of the income distribution. Too much Federal Government is not only inefficient; it also increases the disparity in economic outcomes in the United States.
VI. The Family in Perspective[12]
Before the advent of large scale government welfare, a plethora of private organizations provided assistance for individuals who were economically disadvantaged--churches, aid societies, widows' and orphans' benevolent groups, etc. As Marvin Olasky has documented, these private organizations had an impressive record in reducing poverty and economic distress.[13] They did it largely by relying on human goodness based on values such as love and compassion rather than by mechanistic disbursement of funds. Personal responsibility was required. As Olasky recounts, "No one was allowed to eat and run" at the pre-New Deal private charities. Additionally, families took care of less fortunate relatives. While private organizations such as the Salvation Army continue to help the poor, there is strong evidence that their financial support has been dramatically eroded by the public's knowledge that the government provides public assistance.[14]
When the modern system of public assistance evolved in the 1930s, proponents felt that a humane society should take care of those who had no male breadwinner in the home. No thought was given to the possibility that public assistance predicated on the absence of a male head of household might lead to an increase in the number of such families. If you subsidize something, usually you get more of it, and this has been the case with the single parent families and the welfare system. We believe the evidence supports two propositions: (1) Welfare has increased the incidence of single parent families and contributed to the decline in traditional families (two married parents living together with their children); and (2) Welfare has contributed to illegitimate babies being produced to obtain or increase public assistance payments.
There has been a meteoric increase in the proportion of children not living in two-parent families (Figure 7). Today, about 3 of 10 children live outside the traditional family arrangement, double the proportion of a generation ago.[15] During the same period, real public aid expenditures rose sharply as well. Not only did spending on income maintenance programs rise in real per capita terms, but they rose significantly faster than personal income.
While the literature on the relationship between welfare and family structure was distinctly mixed in the 1970s, more recent studies have more clearly indicated a significant relationship between public assistance and the breakdown of traditional family arrangements.[16] Over 20 years ago, a study showed that AFDC promoted marital dissolution.[17] Several more recent studies have confirmed these notions by revealing significant positive correlations between the presence of single-parent, female-headed families and welfare.[18]
On balance, the evidence is that the programs of the welfare state reduce the need for a traditional marital union, with the division of labor associated with it.[19] Clearly, social scientists and other academics belatedly are discovering what ordinary Americans have known for decades: welfare destroys families.[20] The opportunity costs of getting and staying married have risen as the largesse of the welfare state has provided persons with surrogate spouses, in an economic sense. One consequence of this was described in our report on the impact of the welfare state on children,[21] a persistently high rate of poverty among the one group least able to help themselves, America's children.
A number of significant conclusions have emerged from this analysis of the effects of Federal Government spending on the American family:
First, there is the general descriptive fact that the rate of growth in real median family income has slowed sharply since 1973, amounting to total growth of only 2.5 percent compared to a 103.8 percent rise between 1947 and 1973.The basic thrust of these conclusions is quite the same as those suggested by our earlier studies. At its present size, the American Federal Government is dangerous to our economic health.Second, there is the analytic finding that levels of Federal Government spending that exceed 17.97 percent of Gross Domestic Product produce decreases in real median family income, contributing to explaining the retardation of the post-1973 real income growth.
Third, we conclude that $100 billion of Federal Government spending restraint would increase real median family income by $895.35 (1994 dollars), some 2.3 percent of the 1994 income level.
Fourth, there is substantial disparity by age group in the impact of Federal Government spending in excess of the critical value at which it begins to have negative effects. The younger the householder in a family, the more substantial the impact, both absolutely and relatively. For families with a householder aged 15-24, a decision to restrain Federal spending by $100 billion would increase real income by 10.2 percent.
Fifth, the time pattern of real median family income growth as householders age has undergone a profound transformation. No longer is the assumption of rising incomes from generation to generation as certain as it once was. Our calculations indicate that at least one, and possibly two, generations of Americans have missed out on the American Dream of perpetual generational economic progress, largely because the Federal Government claims too large a share of the nation's Gross Domestic Product.
Sixth, there is no conflict between the twin goals of economic efficiency and economic equity when evaluating the impact of Federal Government spending on the American economy. The level of Federal outlays that is generally consistent with promoting economic efficiency (about 17.5 percent of GDP) also appears to minimize the relative gap between incomes at the top and bottom of the American income distribution.
Conventional wisdom long has held that public policy has worked to reduce income inequality in America, or that, at least, it has the real capability of achieving that goal.[23] For example, the evidence is strong that there was a pronounced decline in measured inequality during the 1930s and 1940s, a period that greatly increased government intervention on both the tax and expenditure side designed to reduce inequality.[24] Increased government expenditures went, or so it was argued, largely to alleviate the plight of the poor and, thus, greater Federal public aid expenditures brought about greater equality. Similarly, it commonly has been maintained that higher and more progressive Federal taxes serve to further reduce inequality in America.
There is another view, however, which holds that government intervention in markets may have a different impact. This has been argued by Tullock, Reynolds and Smolensky, and Bastiat.[25] There are at least five arguments supporting this position. First, there is the important public choice insight that relatively high income rent-seekers may successfully direct public policies toward distributing income to themselves instead of the poor. Second, there are good supply side propositions that suggest that government redistribution efforts might lead to reduced work effort and that leisure-work substitution reaches a point that transfer activities lead increasing numbers of persons to "choose" to be poor.[26] Third, there is the deleterious effect that distributional coalitions can have on growth in income.[27] Fourth, increased public transfer payments may crowd out private charity for the poor. Finally, market adjustments serve to offset many government transfer activities. For example, the farm program may provide "transitional transfers" to farmers, but in the long run the subsidy payments get capitalized in land and other prices and agriculture maintains its long run competitive rate of return (zero economic profits). All of these may combine to nullify the intended redistributive effects of a wide range of government policies.
2. State of the Union Address to the Congress, January 23, 1996.
3. This recent decline has been documented in a series of Joint Economic Committee Reports. See The Middle Class Crunch ... By the Numbers, April, 1996; Reagan Income Growth vs. the Clinton Crunch, March, 1996; Clinton's Middle Class Crunch: Less Income, Higher Taxes, February, 1996; and New Data Confirm Clinton Crunch in Income and Earnings, February, 1996.
4. Changing family size may distort these comparisons somewhat. Since 1973, there has been a significant decline in the number of persons per family, from 3.48 to 3.20 in 1994. If the real median family income estimates are crudely adjusted (by dividing them by the number of persons per family), the percentage increase in real income between 1973 and 1994 is somewhat higher than the 2.5 percent reported here, standing at 11.5 percent. However, there is little impact in the pre-1973 post-World War II years. Thus, the general nature of the pattern of changes in real family income growth is not altered by introducing the family size adjustments. Family size data are from various issues of the Statistical Abstract of the United States.
5. For details, see our The Impact of the Welfare State on the American Worker.
6. Op. cit.
7. Ibid.
8. Similarly, five-year moving averages of these ten-year percentage changes were used in constructing the hypothetical income histories that extend beyond the year 1994.
9. We do not do the age 55-64 to age 65 and over analysis since the age 65 and over category is open-ended, meaning that at any one time it contains people who were in not in the age 55-64 group ten years earlier.
10. Using mean (not median) income data, we calculate that the 1994 mean income of families with householders in the ages 15-44 range was $38,349. For those families with Householders aged 45 and older, it was $42,331.
11. This approach to measuring income inequality is a point in time approach that does not take into account the movement of people between income classes over time. For discussions of the concept of income mobility, see the Joint Economic Committee Minority Staff Studies, Income Mobility and Economic Opportunity, June 1992, and Family Income Growth and Income Equality: Progress or Punishment?, July 1992.
12. This section draws heavily on our The Cost of Waiting for Welfare Reform: A Billion Dollars A Day Doesn't Keep Poverty Away (Lewisville, Texas: Institute for Policy Innovation, October 1994).
13. See Marvin N. Olasky, The Tragedy of American Compassion (Washington, DC: Regnery Gateway, 1992) or his "Beyond the Stingy Welfare States," Policy Review, Fall 1990.
14. Burton Abrams and Mark Schmitz, for example, have estimates that each one dollar of added publicly financed social welfare spending lowers private charitable contributions by 30 cents. See their, "The Crowding-Out Effect of Governmental Transfers on Private Charitable Contributions: Cross-Section Evidence," National Tax Journal, December 1984. Therefore, we would anticipate that reductions in public charity, that is, welfare would produce an increase in private charitable giving for the poor.
15. While this phenomenon is most pronounced among minorities, it is growing quite rapidly among whites as well.
16. For a survey of the earlier literature, see Robert Moffitt, "Incentive Effects of the U. S. Welfare System: A Review," Journal of Economic Literature, March 1992.
17. Marjorie Honig, "AFDC Income, Recipient Rates, and Family Dissolution," Journal of Human Resources, Summer 1974.
18. See Sheldon Danziger, et. al., "Work and Welfare as Determinants of Female Poverty and Household Headship," Quarterly Journal of Economics, August 1982; David Ellwood and Mary Jo Bane, "The Impact of AFDC on Family Structure and Living Arrangements," in Ronald Ehrenburg (ed.), Research in Labor Economics, vol. 7 (Greenwich, Conn.: JAI Press, 1982); and Robert M. Hutchens, George Jakubson, and Saul Schwartz, "AFDC and the Formation of Subfamilies," Journal of Human Resources, Fall 1989. The Hutchens study is somewhat mixed in its findings.
19. See Gary S. Becker, A Treatise on the Family (Cambridge, Mass.: Harvard University Press, 1981) for the standard theoretical treatment.
20. One academic, later during politician, who saw what was happening very early was Daniel Patrick Moynihan. See his The Negro Family: The Case for National Action (Washington, DC: U. S. Department of Labor, March 1965).
21. The Impact of the Welfare State on American Children.
Click here to see Figure 1.
Click here to see Figure 2.
Click here to see Table 1.
Click here to see Table 2.
Click here to see Table 3.
Click here to see Table 4.
Click here to see Figure 3.
Click here to see Figure 4.
Click here to see Table 5.
Click here to see Table 6.
Click here to see Figure 5.
Click here to see Table 7.
Click here to see Figure 6.
Click here to see Table 8.
Click here to see Figure 7.
Return Home