Page images
PDF
EPUB

widely beneficial rather than reflecting only the interests and concerns of a small, already privileged group.

4. On the other hand, if there is a large and growing group in their fifties and sixties who have accumulated little or nothing, this should affect our policies about mandatory retirement age, and suggest the need for ways of supplementing social security more flexible and automatic than the new Independent Retirement Accounts will be. The proposals for a method for converting equity in a home into a cash annuity would also seem more attractive if the number of households who own little else than a home is large.

5. The new Independent Retirement Accounts raise serious regulatory problems, for which we need to know the total financial situation of those who use or plan to use them. The effect of such accounts on capital markets and industrial expansion will depend on the alternative investments available to those who are starting or about to start such accounts.

6. People's information about the 1976 Tax Reform Act, whether they understand it, and their actual or expected adjustments to it can warn us of problems in advance. In particular the combination of Federal estate and gift taxes into a single tax and the raising of deductions to discourage the creation of complex trust forms change many of the rules. Yet, if there is a growing group of affluent households for whom these changes are important, monitoring their effects before they show up in the estate tax returns is obviously important, Official records for some legal forms may be available, but we need to know their relationship with the total household situation, and the extent to which many other households are on the verge of such changes. In spite of overwhelming support by economists for very high estate or inheritance taxes, such taxes have never been very effective in this country. Some information on the attitudes and behavior of people with accumulated wealth may tell us why.

7. Various complex forms of ownership such as trusts have developed and probably increased in numbers. They make official statistics inadequate for assessing changes in wealth distribution. Studies of official records of each type of trust or foundation do not get at their relationships with the whole picture, nor the motivation and purposes of those involved. Only direct questioning of individuals is likely to reveal the way in which sophisticated methods of holding and transmitting wealth are developing and being planned.

8. Since Government policies have differential effects on different kinds of assets and debts, we need data on the distribution of components of net worth, and the extent to which they appear together. Major components such as home equity, savings deposits, common stock, real estate, debts, may differ greatly according to the income or age or total net worth of the households involved. Effects of tax policies, and our evaluation of them will also depend on how the assets were accumulated and why. We know that households differ greatly even at the same income level in their rates and forms of saving, but what is crucial here is the process of saving over extended periods, leading to accumulated wealth. It matters whether the accumulation came largely from savings out of earnings, capital gains, inheritance, or very profitable businesses. Interpretation of wealth distributions and understanding of the system depend on data so that we do not rely on unsupported currently rampant theories about the importance of background, luck, success, or thrift.

It should be apparent from this list of policy concerns and the related data needs, that we need more than the simple dollar facts on the distribution of wealth and its components. We must know how it has been accumulated and by whom and what its owners plan to do with it. Indeed, a focus on precise wealth data is unwise because we probably cannot secure such data without prohibitive cost. One reason there have been no studies of wealth has been a feeling that the data collected would not be sufficiently precise. I think this fear is overdone, particularly because household knowledge of assets has been increased since our earlier methodological studies by the system of reports to IRS and the recipients of the amounts of interest and dividends paid during the year, and by the vesting and information requirements of the recent Pension Reform Act. On the other hand, valuation of assets requires supplemental information from households because market prices have become so unstable. Data on current yield are also needed, since a falling price-earnings ratio is a precise indicator of loss only if the individual sells the stock.

There have been no comprehensive data on the distribution of household wealth since two studies in the early sixties, one by the Survey Research Center, University of Michigan, and the other by the Federal Reserve Board. We at Michigan have collected survey data on some components since then, and the 1972-73

Consumer Expenditure Survey (Census-BLS) contains some asset and debt data in the context of a focus on accounting for income sources and uses.

We have also asked some questions about people's expected retirement benefits which are a special form of wealth and a rapidly changing one. And the Internal Revenue Service has distributional data (by tax returns rather than households) on income from assets-rent, interest, and dividends-which one might inflate to estimated asset values. But such data would exclude unrealized capital gains, funds channeled into Independent Retirement Accounts, accrued private and government pension rights, increased equity in owned homes, and earnings of trust funds. Separate estimates of these other components would not tell us whether they accrued to households with a large amount of other assets. With the many recent changes in tax laws and in the behavior of financial markets and asset values, and with the combination of inflation and unemployment, people face many new situations, and we cannot infer their behavior from the past. We must find out what their experience has been and what they have and will do about it.

Consequently, what is required is interview studies with households, supplemented by information from official sources with proper safeguards to assure privacy, confidentiality, and informed consent. We must have combined information about total household wealth together with information about its sources and expected destinations, and even about the sources of information and advice of wealth holders. Our early study of the affluent indicated that while they got advice, they alone made the decisions about management of their assets, so studies of experts are insufficient.

We are currently collecting some survey data on debts and financial assets for the Board of Governors of the Federal Reserve System, with a focus on the use of credit and the role of Truth in Lending. But those data will not speak to many of the policy issues I have mentioned. Studies of fiduciary returns and asset incomes from personal income tax returns and/or other institutional wealth holders can and should be done, but it is data on the distribution of wealth and its components among individual households that is most needed, together with information on the sources and expected uses of wealth, that will allow informed policy choices.

Mr. FRASER. Thank you very much, Mr. Morgan.

Professor Smith?

STATEMENT OF JAMES SMITH, PROFESSOR OF ECONOMICS,
PENNSYLVANIA STATE UNIVERSITY

Mr. SMITH. Thank you, Mr. Chairman. In spite of the fact that most economic questions really come down to the matter of the distribution of wealth, there is an extreme paucity of wealth distribution information. If one asks a question about the equity of a tax system or of a transfer system, he would like to know something about the various dimensions of economic well-being that are either going to be taxed or supplemented.

In a study of the distribution of wealth in the city of Washington, D.C., that was carried out in 1967, we found that blacks had on average about one-twentieth of the wealth of whites. That was a decade ago, and we should ask what is the situation in the city of Washington, D.C., today? How much change has there been?

We would like to do that study, and probably will carry it out in the near future, but let me go back to that earlier effort and ask a question that just raw measurement begs, and that question is what are the forces responsible for a particular distribution?

One could immediately raise issues of discrimination and stop there, and they would be correct, but how do those discriminatory forces operate to produce roughly 20 times as much wealth to the average white person in Washington, D.C., compared to the average black person. That is a decade after the 1954 decision. Certainly wages and

incomes were much closer between white and black than the ratio of 20 to 1.

One particular cause that comes to mind is the fact that blacks had been systematically excluded from desirable residential markets. They had been forced into acquiring houses in far less desirable areas. Consequently, they did not benefit from secular increase in real estate values that has taken place in this country.

If one asks the question, where will blacks be today when we measure the distribution of wealth in Washington, D.C., a lot will depend on what happened with respect to access to housing. In fact, most wealth accumulated by ordinary folks in this society comes about through the acquisition of real estate, the reduction of the mortgage, and capturing the incremental value of real estate as prices move up. Those who live in Washington, D.C., have experienced substantial increase in house values.

In my comments this morning, I would like to avoid as much as possible, and I think I can completely, discussions of technical details, and of the particular distribution of wealth data that we have in favor of a brief summary. I would like to focus my attention in my oral comments on: (1) this brief summary of what we know about wealth; (2) the sources of wealth information; and (3) the things that can be done to improve wealth information.

I would also like to address this question of why we have so little data on wealth in the country.

Before I get into that, I would like to emphasize that there are some particular uses of wealth information for public policy that the committee and members of legislative branches at all levels of government ought to be concerned with. One has to do with the taxation of income. In the Federal tax structure, we pay a great deal of attention to property income, and we treat it quite differently than we treat income from labor. Particularly preferential treatment that has existed in the past has been eroded over the last decade, but there still remains a preferential favorable treatment of dividend income. There certainly remains favorable treatment of income from municipal securities. There certainly remains a favorable treatment of the deduction of interest paid on residential housing. There certainly remains a favorable treatment of the taxation of capital gains.

With respect to the other side of taxing, that is negative taxing, transfers back to persons who are deemed needful of assistance, the same types of questions arise. You would like to certainly put a floor under people who, by chance or misfortune, find themselves unable to participate very satisfactorily in a society as rich as this one, and the way we have chosen to do this is to transfer income through unemployment systems, social security systems, of various types. In order to make the transfer efficient, in order to get the most benefit to the greatest number of deserving people, we wish to avoid paving benefits to people who "need" the support less than others. That is, we would like to help the worst first.

In order to do this, you need some measure of economic status, and the one that has unfortunately been chosen is income, but yet if we look at data published by, say, the Census Bureau, we find there are a lot of people who have zero or near zero or even negative incomes.

A very large proportion of these people are not poor at all. Many of them are rich. If you have a $5,000 negative income, or $10,000 negative

income, the only way you have been able to afford that type of income is to have been a farmer or businessman who can in 1 year have a big negative income and the next year have a big positive income. Nobody lives year after year with negative income. Those are not poor people but rich people down there. The data overstate the number of people who need assistance.

The practical way to distribute transfers to people at the bottom of the real distribution of well-being is to take account jointly of income and wealth.

Let me move from the use of wealth data in public policy to some discussion now of what we know about the distribution and concentration of wealth, and I would like to note here that most of what we have learned about the distribution of wealth, certainly most of what I have learned, has come about through the support of the National Science Foundation, through the rather energetic support of studies of the distribution of wealth and income by the National Science Foundation, in opposition to a rather enormous effort of the Internal Revenue Service to prevent such studies, and, I might add, some other agencies of Government.

I don't want to leave that remark unqualified, and I will come back to it in the discussion of why we have so little data. Let me say, then, what do we know about the distribution of wealth, particularly the holding of wealth by the very rich?

We have tried to look at wealth distribution over a fairly long period of time, from 1922 to 1972, using a technique which relies upon information that the IRS has begrudgingly supplied, the Federal estate tax returns. The distribution of wealth was particularly concentrated about 1929. The concentration became somewhat more egalitarian as we moved through the Depression and into World War II. Following World War II, we found very little change, no change that we could ascribe to anything other than statistical artifacts.

One then asks, how are things today? Well, they look about like this: If you ask what share of the personally held wealth is owned by the richest 1 percent of the population, then I would respond to you, about one-quarter. If you took even a more rarefied group, the upper one-half of 1 percent, they own about one-fifth of the personally held wealth in the United States. That doesn't include universities or other eleemosynary institutions, just personally held wealth, particularly assets. James Morgan is correct in pointing out that we need to know not only what the total picture looks like, but also about the particular kind of assets that are held, how they are held, and, most importantly, the process by which they are acquired. If we ask what types of assets are held by this top wealth group, this top 1 percent of the population, they own about half of the corporate stock in America. One can say that top 1 percent of the population literally controls all corporate assets in the United States via that half of the value of corporate shares.

Now consider trust assets, a particularly beneficial way for the rich to transfer wealth from one generation to another, the top 1 percent of wealth holders have 91 percent, we estimate, of the total trust assets in the United States, that is, personally held trusts.

The richest 1 percent hold over one-third of all bonds. They own essentially 100 percent of municipals.

In the written testimony I will submit, you can pursue questions about concentration with respect to a few demographic characteristics, age, sex, marital status, and with respect to particular type of assets, but I think the things I have said convey pretty much the picture of the concentration of wealth as it exists in the United States today, which, by the way, isn't as bad as it is in many countries.

The ways that we generate wealth distribution information are basically two. One is the type of source I have been using, Federal estate tax returns. The major problem with the estate tax data is they do not provide us an adequate amount of demographic information to enhance our understanding of the process by which wealth is accumulated.

We have very little demographic information available to us from the estate tax returns, but we get pretty good estimates of the wealth held, particularly at the top.

The other source is surveys, the area where James Morgan has a preemptive position. Surveys have a host of measurement problems, problems that also require a great deal of additional work, and his organization has been pursuing that type of work, and I think it certainly should be carried out.

The ultimate position we would like to arrive at is to get some of the precision of measuring wealth at the top provided by estate tax data combined with the behavioral information that can come from surveys.

We would also like to know a lot more about the process by which wealth moves from one generation to another.

I think the economics profession is probably as much to blame for the lack of knowledge about wealth as any other group. We have in this profession disdained-not James Morgan, but as a profession—-we have disdained getting our hands dirty with measurement. In any other science, if you think about physics, chemistry, biology, an enormous share of its resources go into measurement. In economics we spend a tiny share of the profession's resources in the measurement of economic behavior. Consequently, we find ourselves in a situation where we are always deducing from axioms, what people should do and would do, given certain situations, and, consequently, when we are really faced with tough policy matters, we are usually in a position of not being able to say very much that is empirically grounded.

So I come here somewhat humbly, being a member of that profession, and say to you that we don't know very much-we clearly do not know as much as we should, because we have been doing a bad job of measuring. Now having said that, I can also say that at the Federal level, in the Government, we have done a very bad job of measuring economic behavior and in supporting that kind of work outside of Government.

The 1961 work that was cited, that Dorothy Projector headed up, is a shining example of measurement, but it stands alone. Since 1962, we from measuring anything that you could call important wealth inforin the Federal Government. The census collects a trifle amount of wealth information in the census. It has really shied away from measuring anything that you could call important wealth information. There is the current population survey; there is some work going on in social security, and some going on at HEW, but those are relatively minor efforts. There is also work at the Internal Revenue

« PreviousContinue »