The following testimony was presented on January 31, 2007 before the Joint Economic Committee of the U.S. Congress.
Good morning Senator Schumer and members of the Committee. The Joint Economic Committee (JEC) has just completed 60 years of existence, and during those six decades it has assisted importantly in the making of economic policy, and I am pleased to be part of todays proceedings.
My distinguished colleagues on this panel have painted a somewhat pessimistic and perhaps mildly alarming picture of the American economy. We learn that many Americans have not shared in our nations rising prosperity. The income and wage gap between the rich and the poor is growing. We are told we are becoming a more economically divided nation.
My message is somewhat more optimistic and skeptical of the analysis suggesting that vast portions of the American populace are languishing economically. Let me very briefly touch on three points. First, the conventional measures that are typically cited to denote greater inequality are fundamentally flawed and grossly overstate inequality in this nation, and the growth in it over time. Second, even if one accepts the proposition that America has insufficient equality of economic condition, history tells us that public policy efforts to deal with the problem often are ineffective. Third, some policies that conceivably might lower inequality as conventionally measured would, if adopted, have serious adverse consequences to the economy as a whole.
Turning to the first point, looking at conventional statistics on income distribution, three factors lead us to overstate inequality. First, and probably least important, those statistics are traditionally based on money income, excluding a variety of in-kind, non-cash payments that primarily benefit lower income persons--Medicaid benefits, food stamps, and housing subsidies are three good examples. Any comparison of income levels or income inequality today with, say what existed in 1960 using published income data will tend to overstate any reported rise in inequality, and understate any estimate of income gains for lower income Americans, since non-cash payments have become relatively more important in the intervening time period.
A second factor is that what we should be truly interested in is the economic well-being of Americans, and a far better measure of that economic well-being is consumption spending. Dollar for dollar, people derive more joy from what they spend than from what they earn. As many elementary economics textbooks point out in the first chapter, the ultimate purpose of economic activity is consumption.
We know that in any given year consumer spending is far more equally distributed that income. Comparing the income distribution statistics derived from the Current Population Survey with the BLSs Consumer Expenditure Survey is revealing. For example, the poorest one-fifth year earned only slightly over 7 percent as much income as the richest one-fifth in 2002, but they consumed more than 24 percent as much. Using the most recent data for 2005, we see the richest one-fifth of the population earned 3.47 times as much as the middle quintile, but consumed only 2.31 times as much. Roughly speaking, conventional measures show consumption inequality is at least one-third less than for income inequality.
The third point relating to the overstatement of inequality relates to the remarkable income mobility of the American people. For example, at the request of this Committee, the Treasury Department in the 1990s provided data suggesting that the overwhelming majority of persons in the bottom quintile of the income distribution were in another quintile a decade later, and a large percent even moved up or down the distribution from one year to the next. Researchers at the Urban Institute and other organizations have made similar observations. This phenomenon helps explain the narrowness of the distribution of consumption spending relative to the distribution of income, as observed decades ago by the late Milton Friedman and in a different context by Albert Ando and Franco Modigliani. Failure to consider the income mobility of people contributes to the inadequacies of traditional measures of income distribution and also leads us to create some inequities and inefficiencies when devising tax policies based on single year definitions of income.
While we are talking about measurement problems, they are particularly prevalent in our discussions of changes in earnings over time. Go to page 338 of the 2006 Economic Report of the President. We learn that average weekly earnings of workers in private nonagricultural industries in 2005 were over eight percent less than they were in 1964, the year Lyndon Johnson announced his Great Society initiatives. Yet turn the page, to page 340. Looking at real compensation per hour in the non-farm business sector for the same time period, we learn it has risen 75 percent. Page 338 is consistent with a Marxian or even Malthusian interpretation of the economy--a tendency for wages to fall to near subsistence, and evidence of mass exploitation of the working proletariat by exploitive capitalists. Page 340 is consistent with the view that with economic growth, the earnings of workers have risen sharply, and also consistent with national income accounts data that shows per capita real consumption has increased about two percent annually.
Yet even the data on page 340 suffer from deficiencies. We learn that productivity per hour in the non-farm business sector in 2005 was 2.28 times as great as in 1964, yet compensation rose only 1.75 times, a pretty big difference that is inconsistent with the neoclassical economic theory of factor prices and suggestive that owners of capital are indeed deriving extraordinary profits as a result of paying workers less than what they contribute to output at the margin. This should have resulted in a significant decline in compensation of workers as a percent of national income. Yet the national income data taken from pages 314 and 315 of the same source show a radically different story. Compensation of employees actually rose from 60.75 to 61.51 percent as a percent of the national income. The share of national income accounted for by corporate profits fell slightly in the same time period.
I am making two points here. First, interpretations of economic data can be exceedingly misleading. Second, the analysis of broader measures of economic performance suggests that workers as a group have shared in our national prosperity of the past several generations. The original wage data I cited suffer from two enormous deficiencies. First, they fail to take account non-wage forms of compensation, particularly health care and retirement benefits. These have soared in magnitude over time. Second, the calculation of changing values in constant dollars is fraught with peril, and the Consumer Price Index used in these calculations very significantly overstates inflation in the eyes of virtually every mainstream economist, liberal, conservative, vegetarian, Presbyterian, what have you. Similarly, analysis of wage changes by wage or income category suffers not only from these problems, but from the aforementioned phenomenon of the rapidly changing economic status of individual members of our opportunity society over time.
You dont need a Ph.D. in economics to observe that never has a society had a middle class more used to what once were considered goods and services available only to the uber rich. Middle income Americans live in larger homes, buy more gadgets like iPods and cell phones, live longer, are more if not better educated, and take nicer vacations than either their parents did or do and their counterparts in any other major nation. I returned two days ago from a two week cruise in the Caribbean, traveling less with top business executives or even elite Ivy League professors than with equipment salesmen, butchers, and teachers--ordinary folk. That simply did not happen even 30 years ago.
My second major point relates to public policy dealing with economic inequality. Time does not permit a detailed exegesis of past efforts. But a reminder of some historical experiences is sobering. Policy can come from the tax, spending or regulatory side. I will ignore regulatory matters in the interest of time, although I would hasten to commend Senator Schumer for recent statements showing his concerns about the abusive use of the tort system as a growth-impeding way of redistributing income. Looking at taxes, attempts to make the system more progressive often have unintended effects. For example, sharp reductions in top marginal tax rates in the 1920s, 1960s, and 1980s, viewed by some as favoring the rich, actually led to sharp increases in the tax burden of the rich relative to the poor. I worked for this Committee during the 97th Congress in 1981 and 1982 in a political environment much like today with divided government, with the Republicans controlling the Executive while Congress was more under Democratic control, yet the two branches managed to work together to fashion a more growth oriented tax policy with lower marginal tax rates that contributed mightily to the boom that has followed. I hope the 110th Congress is capable of similar accomplishments.
Taxes have behavioral consequences. The CBO greatly underestimated revenues that would arise from the reducing in the top capital gains rate to 15 percent, for example. Falling rates unlocked billions in unrealized gains that have helped fund our rapidly expanding government. Similarly, sharp reductions in the number of estates subject to death taxation as a result of reform in those laws has not led to a sharp decline in revenues from that source, as some had expected. It would be a tragedy to reverse the positive effects of the tax reductions of the past few years that, like the Kennedy tax reductions of the 1960s, have had a positive impact on economic activity.
On the spending side, history again shows disappointing results of many initiatives to help the poor or middle class. As the January 20 issue of the Economist notes, government job training programs have internationally been largely failures. Spending initiatives in the areas of education, medical care, and public assistance have usually brought about disappointing results. Despite spending far more in real terms per student than a generation or two ago, American students do not appear to be learning much more, and the education for lower income students is particularly deficient. A tripling of federal aid to college students since 1994 has been accompanied by a decline, not an increase, in the proportion of students from the lowest quartile of the income distribution attending and graduating from our finest universities, which are increasingly becoming taxpayer subsidized country clubs for the children of the affluent. While Medicaid has brought some increase in medical care for the poor, it has done so at an enormous cost to society, and the cost pressures of a highly inefficient system are leading companies to cut back on health care benefits for working middle class Americans. As to public assistance, it is far greater today in real per capita or per poor person terms than in 1973, yet the current poverty rate is higher. The welfare reforms of the 1990s were an important achievement, but the overall picture is, at the very least, mixed.
Speaking of public assistance, I have to make one statement that may sound a bit callous or insensitive to some, but it is an important but often neglected truism. Comparing the rich and the poor, it is worth noting that the rich work a lot more. Of those persons in poverty, only a tiny minority work full-time. We have relatively few working poor in America. And it is worth noting that employment creation is greatest in periods when the government allows the incredible job machine generated by the competitive private sector operating in a market environment to work. The job creation of the 1980s was stimulated by a halt to the growth in governments share of GDP characterizing earlier decades, and by tax reductions that stimulated the spirit of enterprise. The job creation of the 1990s was stimulated by an unprecedented decline in government expenditures as a percent of GDP for eight consecutive years--a reverse crowding out phenomenon that propelled an enormous outpouring of American creative and entrepreneurial endeavor.
Turning to my final point today, there is a temptation to do things in the interest of protecting middle and lower income Americans that might have highly undesirable effects on the economy as a whole. In this regard, the rise in protectionist sentiment in Congress is appalling, particularly as is largely centered in a party which historically has favored free trade, a policy that has brought prosperity to almost all Americans while at the same time has contributed enormously to eliminating global disparities in the distribution of income and wealth. I hope the intelligent wing of the Democratic Party, represented by able persons such as those who preceded me on this panel, will be able to prevent a return to policies reminiscent of that old Democratic bete noire, Herbert Hoover. The Smoot-Hawley Tariff and rising taxes were a factor, along with Hoovers inane wage policies, for the Great Depression of the 1930s. Let us not repeat that today. I hope the Democratic Party will try to emulate Franklin D. Roosevelt, John F. Kennedy and Bill Clinton in the area of trade policy, not Herbert Hoover.
At a macro level, I believe the biggest single factor in the modest slowdown in growth rates in this decade relative to the 1980s and 1990s is the sharp increase in government expenditures. From fiscal year 2001 to fiscal year 2006, total federal outlays rose by 42.4 percent, or $790.1 billion. By the way, the overwhelming majority of that was for non-defense or national security purposes. This was nearly double the percent growth in GDP. Receipts rose well over 20 percent or roughly equal to the growth in GDP, so the burgeoning deficit reflected a spending binge that resulted in some crowding out of private economic initiatives. Dollar for dollar, the evidence is crystal clear that private spending has more productivity-enhancing effects than public spending because of the discipline that competitive markets impose on market enterprise. The tax cuts largely corrected for the natural tendency for taxes to rise relative to national output. Raising taxes again would reduce the deficit, but would have direct unfortunate disincentive effects on human economic behavior and would also reduce the political costs to Congress of incremental spending initiatives, which almost certainly would have severe economic effects. I hope some early indications of spending constraint are maintained in the months and years ahead. While I am not the financial guru that Secretary Rubin is, an analysis that I have conducted with Lowell Gallaway for this Committee in the past suggests that the two best determinants of the growth of wealth as measured in equity prices are the rate of inflation and government spending as a percent of GDP. Rising government spending is associated with falling market values and wealth, with all the adverse consequences that has for pensions. And stable prices are much better than inflation. The Fed has done a pretty good job on the inflationary front, but the Congress and the Executive are guilty of having shown insufficient constraint with respect to federal expenditures.
Again, I praise the JEC for providing a needed forum for the analysis of policy possibilities informed by factual evidence. I hope the next 60 years are as successful for this Committee as the last 60 have been.