This is the third JEC study in a series I have commissioned on the impact of the welfare state on various aspects of the American economy and society. The first study, The Impact of the Welfare State on the American Economy, examines the drag on economic growth resulting from excessive levels of federal spending. The second study, The Impact of the Welfare State on Workers, analyzes the relationship between the size of the federal government and recent trends in income and compensation.
This third JEC study examines the impact of the American welfare state on poverty, especially child poverty. It also analyzes the effects of excessive federal intervention on other aspects of childhood and family life. The study concludes that excessive federal spending and intervention are increasing overall poverty as well as child poverty. Excessive federal spending and intervention are also related to a broad deterioration in a variety of economic and social factors affecting children and families. As in the first two JEC studies in this series, the central problem identified is that government spending and intervention have reached a scale that is immoderate and counterproductive.
Among the conclusions in the new JEC study are:
The excesses of the modern welfare state are one cause of the deterioration in a number of economic and social factors affecting children and families. The increases in family breakdown, illegitimacy, educational failure and other social pathologies are related to the perverse incentives of the excessive and impersonal welfare state, and its counterproductive effects on economic and income growth and poverty.
I am pleased to make this study available to contribute to a thoughtful debate on poverty and welfare reform issues.
Imagine America's 70,000,000 children marching by in a steady procession. As they pass, you can count on one thing; on average, one of every ten will be classified by the Federal government as officially living in poverty precisely because government is too large. That is the major finding of this, our third, study of the question of the impact of the welfare state on American economic life. To explain the rationale for this conclusion, we begin with a technical analysis of recent trends in child poverty in the United States.
Of all the statistical data reported by the Federal government, probably the most heart-wrenching are the child poverty numbers. Figure 1 shows the behavior of this data set beginning with 1959. Clearly, there are distinctly different periods in these data. Between 1959 and 1969, the child poverty rate falls precipitously, from 27.3 to 14.0 percent, an average of 1.33 percentage points a year. After 1969, however, the official child poverty rate begins to drift upward, reaching a new peak in 1983, at 22.3 percent. It then moves downward to 19.5 percent in 1988 before rising to 22.7 percent in 1993. In the 24 years between 1969 and 1993, the child poverty rate rises by 8.7 percentage points, about 0.36 percentage points a year.
Why did we have this variable pattern of behavior of the child poverty rate? Understanding it requires a grasp of the factors that influence more general poverty rates, such as that for families. Poverty among children is derived largely from the broader phenomenon of adult poverty. Also shown in Figure 1 is the behavior of the family poverty rate. It closely follows the child poverty rate, at first declining, this time until 1973, then moving upward until 1983, dipping through 1989, and, finally, rising by about twenty percent (two full percentage points) by 1993.
The close relationship between the child and family poverty rates is shown in the scatter diagram of Figure 2. This suggests that a good starting point for accounting for variations in child poverty is the development of an explanation for movements in the family poverty rate. For this purpose, we focus on three broad factors:
As a measure of the first of these, we will use the data for real compensation of workers developed in the second report in this series.[1] They are shown, beginning with the year 1959, in Table 1. Also included in Table 1 are data detailing the behavior of the unemployment rate and the percentage of all money income received by the lowest quintile of income recipients. The unemployment rate will be used to control for the differences over time in the stability of the flow of labor market income to families.
What about the percentage of income received by those in the lowest income quintile variable? Its role is multiple. First, it takes into account the nature of the distribution of labor market income. In addition, it also reflects the availability of other types of income to those in the bottom quintile of the income distribution. Of particular importance in this regard is the effect of government programs aimed at redistributing money income to the poor. On the one hand, many government programs provide money income transfers that have the "direct" effect of enhancing the level of money income of families. However, there is an accumulation of evidence that indicates that these transfers may also produce changes in labor force behavior, changes leading to a reduction in money income generated by working.[2] The combined impact of these effects will be reflected in the percentage share of money income received by those in the bottom quintile of the income distribution.
When we use the data found in Table 1 to explore statistically the variations in the family poverty rate, we obtain the results shown in Table 2. Collectively, these three variables explain 98 percent of the variation in the family poverty rate and all of them are highly significant in a statistical sense. In addition, the direction of the relationships is the expected one in all cases. Higher levels of labor force income reduce poverty, a higher unemployment rate increases it, and a larger share of income received by those in the lowest quintile of the income distribution leads to a reduction in poverty.
The statistical relationships reported in Table 2 permit the calculation of "predicted" values for the family poverty rate for the years in question. These predicted values then can be used to explain variations in the child poverty rate. From earlier research we have conducted dealing with child poverty, we learned that the child poverty rate is more sensitive than the family poverty rate to the underlying economic variables.[3] Consequently, we experimented with including the unemployment, percent of income received by the lowest quintile, and worker compensation variables, along with the predicted values for the family poverty rate, in a statistical explanation of the child poverty rate. Two of these variables, the unemployment and percent of income received by the lowest quintile measures, are statistically significant factors in explaining movements in the child poverty rate. The results of the statistical analysis are shown in Table 3. The three explanatory variables account for 96 percent of the variation in the child poverty rate in the United States.
The statistical patterns shown in Tables 2 and 3 set the stage for determining the impact of the magnitude of Federal government spending on child poverty. For example, from the second report of this series, we know that the real compensation of workers has been influenced significantly by the size of the Federal government. In that study, we found that Federal spending in excess of 17.4 percent of Gross Domestic Product exerts a negative impact on the productivity of labor and real compensation of workers. In 1994, real compensation of workers would have been 13 percent higher were it not for excessive Federal government spending. On the basis of these findings, we can calculate the effect of government spending on the child poverty rate because of its impact on the level of work related income that is typical for the economy.[4] Based on fiscal year 1994 Federal government spending amounting to 22.0 percent of Gross Domestic Product, we estimate that this dimension of the impact of excessive Federal government spending accounts for 2.55 percentage points of the child poverty rate.
What about the other factors influencing the child poverty rate? Are they, too, affected by the size of Federal government spending? For example, is there a systematic relationship between the volume of Federal government spending and the percentage of all money income received by those in the bottom quintile of the income distribution. We have explored this question in a statistical fashion with one significant modification of our earlier analysis. Rather than relating the low quintile income share to all Federal spending, we chose to focus on just non-defense outlays. This is the area of the Federal budget that includes the various programs specifically designed to affect the distribution of money income in the United States. As in our earlier studies, we examined the statistical properties of the following relationship:
where L denotes the share of all money income received by the bottom quintile of the income distribution and Gn represents non-defense Federal government spending expressed as a percentage of Gross Domestic Product.
The statistical results (shown in Table 4) are similar to those described in our previous studies. Both forms of the Federal government spending variable show a highly significant statistical relationship with the income share of the bottom quintile, but in opposite directions. Up to a certain level, non-defense Federal government outlays have the effect of increasing the share of all money income that goes to those at the bottom of the income distribution. However, beyond that threshold, the impact of further non-defense spending on the low income share is negative, pulling it down. The critical level of non-defense spending is 13.1 percent of Gross Domestic Product. That level of spending, combined with the 4.2 percent of Gross Domestic Product that was spent for defense purposes in fiscal 1994, is consistent with total spending of 17.3 percent of Gross Domestic Product, very nearly the same as the 17.4 percent spending level that we found to be optimal in the productivity and compensation analysis and the 17.6 percent observed when dealing with the impact of Federal spending on Gross Domestic Product itself.
The hard reality is that non-defense outlays in fiscal 1994 amounted to 17.8 percent of Gross Domestic Product, over a third more than what would be optimal from the standpoint of maximizing the income share of the bottom quintile and minimizing the negative effect of inequality in the income distribution on child poverty. We calculate that excess Federal non-defense spending (the difference between 17.8 and 13.1 percent of Gross Domestic Product) has the indirect effect, operating through its impact on the family poverty rate, of increasing the child poverty rate by 3.48 percentage points. In addition, there is a direct effect on the child poverty rate from this source of another 3.05 percentage points.[5] Thus, the reduction in the share of money income received by those in the bottom quintile of the income distribution that can be attributed to excessive Federal non-defense spending leads to a cumulative increase in the child poverty rate of 6.53 percentage points.
Using this figure, we calculate that the indirect effect of excessive Federal spending on the child poverty rate, induced through the impact of unemployment on the family poverty rate, is to increase it another 0.43 percentage points. The direct effect is an additional increment of 0.36 percentage points. These are minimal numbers due to our using the 17.5 percent figure rather than the 16.1.
The various effects on the child poverty rate attributable to excessive federal spending, as of fiscal year 1994, are summarized in Table 5. Their total impact is to increase the child poverty rate by 9.87 percentage points. The meaning of this was pointed out in the opening sentence of this report: On average, one out of every ten American children lives in official poverty precisely because of excessive levels of Federal government spending.
To illustrate the importance of additions to Federal outlays at present spending levels, we have calculated the impact on child poverty of restraining Federal non-defense outlays by $100 billion. That would reduce the percent of Gross Domestic Product claimed by Federal spending by about 1.5 percentage points. The impact on the number of children officially classified as poor would be profound. The child poverty rate would fall by 4.35 percentage points, moving over three million children out of official poverty. Put another way, every $33,000 of non-defense spending restraint would move a child out of poverty. Probably the most effective anti-child poverty program we can have is one that embraces fiscal restraint at the level of the Federal government.
Having completed the description of the more technical aspects of our analysis, we turn now to fleshing out the story of the Federal government's effect on the economic and social status of American children, focusing on the human dimension of the issue. While the increased size of government in modern times has accentuated child poverty in the United States, the quantity and quality of life for children has suffered in other ways, in part because of the pernicious effects of governmental intervention. Consider the eight following stylized facts:
The quantity of children is the smallest in relation to the total population as at any time in American history (see Figure 3). In part, this reflects the rising life expectancy of Americans, but to a considerable extent it is a product of falling birth rates. Birth rates in the U.S. are currently only slightly over 15 per 1,000, compared with more than 25 per 1,000 in the mid 1950s.
While part of the reduction in the presence of children in our lives stems from factors associated with economic growth and urbanization, government policies have contributed to this reduction in child quantity. The real dollar amount of dependent deductions on federal income tax returns is far lower today than in, say, 1950. Removal of governmentally enforced bans on abortions and the subsidization of birth control programs in schools and through public assistance have served to reduce births. A little known phenomenon is that the welfare system apparently powerfully encourages abortion.[6]
A growing proportion of American children do not have the support and nurturing that comes from the traditional two parent family. The number of illegitimate births more than tripled from 1970 to the early 1990s, exceeding 30 percent of the total in recent years. Whereas, as late as 1970, only 13 percent of families with children had only one parent, by 1994 the proportion had more than doubled, to 31 percent. Over 15 million children live in family situations where no father is present. From 1960 to 1980, the divorce rate (divorces per 1,000 population) increased by over 100 percent; while modest decline has occurred since 1980, the rate is still more than twice 1960 levels. The net marriage rate (marriages minus divorces, per 1,000 population) is now at the lowest level in modern history. Moreover, because of increases in female labor force participation and second job/overtime work, the amount of time that parents spend with their children is declining even in traditional families.
There is voluminous evidence that governmental policies have contributed to the breakdown in traditional loving family arrangements for children. Four examples will make the point. First, public assistance programs have provided powerful incentives for illegitimacy and for the existence of single parent families.[7] Second, no-fault divorce laws at the state level have reduced the legal and psychic costs of getting divorced. Third, the still-existing "marriage penalty" reduces the federal tax burden for some couples who live together outside marriage. Fourth, the same tax code provides tax breaks for families who hire outsiders to care for their children, but not for family members who provide such care.
There is general agreement that skills inculcated in younger people increase their productivity and earnings potential later in life. A huge body of literature suggests that despite increases in resources used to finance education, learning by American youth is poor relative to America's past or to other countries.[8] While real public spending per pupil has risen sharply in the U.S. and is above that for almost all industrialized nations, Americans tend to rank below average in international test comparisons. SAT scores are lower now than a generation ago, and performance on other tests (e.g., the Iowa Test of Basic Skills) shows little change, reversing a historic pattern of improvements in learning over time.[9]
Research results indicate that the previously mentioned breakup of families and the growth in the welfare population has contributed to the problem of mediocre academic performance.[10] Yet there is a strong feeling that another significant factor relates to an inefficient, costly system of educational delivery. Most Americans have little or no choice of schools, with educational services provided by a monopolistic government provider who has little incentive to improve educational quality, respond to parental and community needs, and so forth. The rise in strong labor unions over the past generation has increased the use of children as pawns in efforts to increase salaries and benefits for employees.
The long term consequences of this are clearly negative for today's children. John Bishop has suggested that the productivity decline and educational quality decline in America are intimately related.[11] The worldwide decline in the relative compensation of unskilled workers relative to those with greater "human capital" suggests that our children could suffer significantly in an economic sense from the lack of willingness to enact meaningful reforms in educational delivery.[12] While such ideas as vouchers, charter schools, alternative certification for teachers, repealing teacher tenure and so forth may have promise, powerful political forces (particularly teacher unions) have generally limited implementation of reform. Recent efforts in Congress to prevent even modest experimentation with school choice in the District of Columbia schools, despite support by the public school's top administrator, illustrate the problem.
The ultimate expression of disenchantment with life is suicide. During the 1960s and 1970s, there was a huge increase in suicides among younger Americans. For example, from 1960 to 1977, the suicide rate (per 100,000) of white males aged 15 to 24 more than doubled (going from 8.6 to 22.9). The problem continues to worsen. From 1980 to 1992, the suicide rate for all 15 to 19 year old Americans rose from 8.5 to 10.8, and the rate more than doubled for younger (10-14 year old) children. The rate nearly tripled for black males in their late teens. Meanwhile, the non-child suicide rate for Americans was falling; for example, the rate for individuals from 45 to 54 years of age fell from 15.9 to 14.7.
Clearly, despair among children in America has sharply increased in the past third of a generation. An increasing proportion of kids feel unwanted, unloved. Policies that keep parents away from children, or that have contributed to crime and drug use among children, are at least partly to blame.
The economic prosperity of our children and grandchildren depends on their having a large stock of physical and human capital to employ, providing a high productivity that allows for high living standards. Aside from the erosion in human capital formation stemming from poor educational performance, the rate of savings and investment in America has fallen in recent years from earlier American norms and standards in other nations.
Capital formation - investment - is financed through savings. In the 1960s, 1970s, and early 1980s, Americans typically saved 7 or 8 percent of their disposable income. In the 1991-95 period, however, the median annual savings rate was a paltry 4.5 percent. While the savings rate in 1994 in the U.S. was 4.1 percent, it ranked from 10.4 to 16.8 percent in Great Britain, Germany, France, Japan and Italy. Our neighbors in Canada saved at nearly double our rate, 7.6 percent. A broader definition of savings would include corporate and business savings (retained earnings) as well as government savings (or dissaving). In the first half of the 1960s, in a typical year gross savings in the U.S. equalled over 21.7 percent of gross domestic product, compared with under 15 percent in the median year in the 1990-1994 period. To supplement limited domestic savings, Americans are importing savings from other nations, leading us to becoming the world's largest debtor nation.
Federal tax policy is virulently anti-savings. Capital gains taxes often are confiscatory, sometimes imposing levies on completely fictitious gains. Suppose a 40 year old woman bought $10,000 in stock in ABC Corporation in 1970. Suppose at the beginning of 1996, she sells the stock for $30,000, as she needs to reinvest in higher yielding securities to provide retirement income. The government claims that the woman has a $20,000 "capital gain" on which she will typically pay about $5,600 in federal income taxes (and often more to state and/or local governments). Yet the purchasing power of the $30,000 in 1996 is less than the $10,000 in 1970. In real terms, the stock has lost value, yet the government imposes a substantial tax on inflation-induced fictitious income.
Other anti-savings taxation provisions abound. Corporate profits are taxed. Dividends passed on to taxpayers are taxed again. Retained corporate earnings are taxed again at the individual taxpayer level through capital gains levies. Estate taxes impose still other levies on savings. America's failure to save is easy to explain - the government has made it costly to do so. A smaller savings pool pushes up interest rates and reduces capital formation. The real losers are our children who will grow up with a smaller capital stock to work with, lowering their standard of living.
It is probable that the gross savings rate of Americans has been lowered by negative government savings in the form of deficit financing of expenditures.[13] In the first 200 years of the Republic under the Constitution, the federal government ran 107 budget surpluses and 93 budget deficits. In almost half the years, the Federal government was able to pay its bills only by borrowing, an act that inflicts a cost on future generations. Things have worsened in very recent times. This is the 27th consecutive year in which the federal government has resorted to deficit financing; most have been years of peace and prosperity.
The rise in deficit fiscal behavior reflects the breakdown in an unwritten rule (our fiscal constitution, if you will), that said that borrowing is bad and should be reserved for economic and military emergency situations. This breakdown began with the economics of John Maynard Keynes.[14] Political entrepreneurs have learned that votes are maximized by engaging in debt financed spending. Efforts to correct this imbalance, such as a balanced budget amendment to the constitution, have been thwarted by special interest groups and liberal politicians unwilling to accept constraints on their fiscal authority. The inability to arrive at a seven year deficit elimination plan under current institutional arrangements increases the arguments for greater restraints on fiscal behavior that is inimical to the welfare of future generations of Americans.
Statistics show that criminal activity has soared among our youth. In part, this criminal activity has involved the consumption and distribution of dangerously addictive illegal drugs such as cocaine and heroin. The number of juvenile arrests for violent crimes more than doubled from 1970 to 1993. A majority of individuals arrested in 1993 for arson were teen-agers or younger. Arrests for the sale or manufacture of heroin and cocaine rose 14 fold for teenagers from 1980 to 1993.
Part of the rise in destructive criminal behavior among our youth may result from excessive permissiveness in our system of criminal justice, serving to make crime pay where previously it did not. Another part, however, arises from a decline in a sense of moral absolutes in the population. Increasingly, Americans do not know the difference between right and wrong.
Again, this increased moral relativism is at least in part a result of governmental policy. Court rulings prohibiting the posting of the Ten Commandments or the use of prayers at graduation ceremonies are small examples demonstrating a downplaying of institutional forces that inculcate values. The welfare system undermines the negative consequences of behavior historically considered "bad", such as illegitimacy. The schools have promoted curricula that reduce positive reinforcement of traditional values.
In short, government has served to intervene in the creation of healthy human relationships that develop in the state of nature and exist in virtually every known human society. Aside from impoverishing children financially, government policies probably, on balance, have had debilitating effects on the future moral state of our progeny.
The record of government involvement in the lives of our children is a sad one. Through a wide range of public programs, children have been impoverished in large numbers, both in a purely financial sense and in the conditions of their daily life.
Just the pure size of government, especially at the Federal level, is imposing immense burdens on young people, shifting approximately one in every ten into a life of official poverty. Accompanying this poverty status is an erosion of the very nature of family life in America, alienating large numbers of children from the historic economic and social norms that have been typical of the United States. All this has been accomplished, for the most part, in the name of "helping" our children.
Perhaps no single group in American society better understands the principle that the outcomes of public policy initiatives almost always involve "unintended consequences." Deprived of economic wherewithal by government policies that encourage their parents to reduce their economic activity, deprived of moral leadership by public policies that embrace the concept of moral relativism, too many children have been left to drift, often into lives of crime and worse. To them, there is perhaps no greater mockery than the canard, "I'm from the government and I'm here to help you."
Lowell E. Gallaway is research fellow at the Independent Institute and distinguished professor of economics, Ohio University. He received his Ph.D. in economics from Ohio State University. He has been staff economist, Joint Economic Committee of the Congress of the U.S.; chief, Analytic Studies Section, Social Security Administration; and has taught at the Colorado State University, Lund University, University of Minnesota, University of New South Wales, University of North Carolina, University of Pennsylvania and University of Texas.
Richard K. Vedder is research fellow at the Independent Institute and distinguished professor of economics and faculty associate, Contemporary History Institute, Ohio University. He received his Ph.D. in economics from the University of Illinois. He has taught at the University of Colorado, Claremont Men's College, and MARA Institute of Technology.
Endnotes
1. The Impact of the Welfare State on Workers, Joint Economic Committee of the Congress, March 1996.
2. For some detailed evidence of this phenomenon at work, see Lowell Gallaway, Richard Vedder, and Robert Lawson, "Why People Work: An Examination of Interstate Variations in Labor Force Participation," Journal of Labor Research, Winter 1991, pp. 47-59. This analysis reveals that the proportion of people in a state receiving public aid is positively related to the attractiveness of welfare benefits relative to wages and that the labor force participation rate among states is smaller the larger the proportion of people receiving public aid. For further confirmation of these linkages, see M. Anne Hill and June O'Neill, Underclass Behaviors in the United States: Measurement and Analysis of Determinants (New York: City University of New York Baruch College, August 1993) and Robert Moffitt, "Incentive Effects of the U. S. Welfare System: A Review," Journal of Economic Literature, March 1992.
3. Lowell Gallaway and Richard Vedder, Poverty, Income Distribution, the Family and Public Policy (Washington, DC: Government Printing Office, 1986), Chapter VII.
4. This is done in a two-stage fashion. First, the impact of depressed levels of compensation of workers on the family poverty rate is estimated. Then, this effect is translated into a change in the child poverty rate. In both calculations, the regression coefficients reported in Tables 2 and 3 are the basis of the estimates.
5. The indirect effect was estimated in the same fashion used in calculating the impact of the compensation of workers variable on the child poverty rate. The direct effect is measured in a simple fashion by the coefficient associated with the low quintile income share variable in the child poverty rate regression (Table 3).
6. For details, see Lowell Gallaway and Richard Vedder, "Unintended Consequences of Welfare Reform?", IPI Insights, June/July 1995, pp. 1-2.
7. See Robert D. Plotnick, "Welfare and Out-of-Wedlock Childbearing: Evidence from the 1980s," Journal of Marriage and the Family, August 1990; C. R. Winegarden, "AFDC and Illegitimacy Ratios: A Vector Autoregressive Model," Applied Economics, March 1988, and M. Anne Hill and June O'Neill, "Underclass Behaviors in the United States ...," op. cit.
8. See, for example, Eric A. Hanushek, et. al., Making Schools Work (Washington, DC: The Brookings Institution, 1994); and Diane Ravitch and Chester Finn, What Do Our 17-Year Olds Know? (New York: Harper and Row, 1987).
9. The mean SAT score fell from 958 in 1967 to 912 in 1994. See the 1995 Statistical Abstract of the United States, p. 175. Some evidence from the National Assessment of Educational Progress suggests math and science proficiency is improving somewhat over time, but not reading and writing performance. Ibid., p. 176.
10. The most definitive study is John E. Chubb and Terry M. Moe. See their Politics, Markets & America's Schools (Washington, DC: Brookings Institution, 1990).
11. John Bishop, "Is the test Score Decline Responsible for the Productivity Growth Decline?" American Economic Review, March 1989.
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 Table 5.
Click here to see Figure 3.
Return Home