dojN002032.xml

dojN002032.xml

born-digital

email

yes

no

2002-01-18

Friday, January 18, 2002 3:42 PM
Comments on Interim Rules





Attachment 1:


Dear Mr. Feinberg:

You and I have never met, but we have both worked with many of the same lawyers. From what I have learned, you are an excellent lawyer. Prior to your appointment, I had the privilege of meeting with many of the lawyers at the Department of Justice regarding the issues faced by those administering the compensation fund. Please accept the following comments as referring solely to your economic advisor(s). While I do not question their motives, their advice and analysis is simply wrong when judged against the economic literature on human capital and economic loss.

The interim rules have been subjected to severe criticism as being "unfair". Generally, these criticisms are either that, as an alternative to litigation, they undercompensate families or that, as a social program, they are unfair because they do not provide equal compensation to all claimants. Arguably, if the liability test cannot be met or, if it is met, there is no money available from the airlines, any money from the fund is more than families would otherwise receive. This notion, however, constitutes the proverbial slippery slope with political implications. Consequently, the following comments are offered as if your goal is to provide compensation for the full financial loss to a family or, in the case of a single person without dependents, to his or her estate.

It is understood that you have attempted to be "fair", and that you have endeavored to hold down the cost to taxpayers. However, it is also understood that you are a lawyer, not an economist. Furthermore, from the rules, it appears that your input on economic-related issues has come from those who have established themselves as "experts" in the courtroom, where the only test is that they can convince a group of laymen, including the lawyers, that they know something about which they speak. Unfortunately, these "experts" are almost universally unfamiliar with the economics of human capital, the very issue to be addressed by you and the fund. That is, it appears that your advice has come from the typical courtroom economist who, while quite able to convince a jury of his or her "expertise", generally lacks education in the very part of economic science underlying the issues being addressed.

Universally, your rules reflect the approaches taken by so-called litigation economists, many of whom belong to the relatively new association of plaintiff "experts" known as the National Association of Forensic Economics (NAFE). The journal of that association is the Journal of Forensic Economics. This journal is the sole reference for most of the citations these economists make to the "scientific literature". Unfortunately, this journal is not included in research libraries of universities granting a Ph.D. in economics. It is only available in a few law libraries for the obvious reasons. Noticeably lacking are articles in that journal by authors from the ranks of legitimate human capital economics. For example, we have never seen an article published in that journal by an author who has also published in the American Economic Review, the flagship journal of the economics profession. Certainly no Nobel prize winner, such as Gary Becker and James Heckman, who received the Nobel prize for their work on human capital economics, has ever published in that journal. Moreover, we have found no citation to an article in that journal by any article in the American Economic Review.

According to a newspaper report, NAFE met "jointly with the American Economic Association" (AEA) during its recent annual meetings in Atlanta. This seems to provide the imprimatur of the AEA to NAFE activities. In fact, there are over fifty associations who are loosely affiliated with the Allied Social Sciences Associations (ASSA). ASSA associations hold their annual meetings at the same time and place each year. The fact that NAFE and AEA meet annually as part of the ASSA is pure happenstance. No one at the AEA administrative offices, nor the president of the association, was aware of this misrepresentation. Moreover, the so-called "joint panel to discuss problems with the compensation fund" chaired by the joint editors of the NAFE journal, was a "press conference" called by NAFE. According to the AEA, it had nothing to do with NAFE's press conference. Clearly, the American Economic Association, whose members include virtually all American winners of the Nobel Prize for economics, all winners of the John Bates Clark Medal, all members of the President's Council of Economic Advisors, is not going to participate in a joint panel chaired by the editors of a special interest "journal".

The following comments come from a human capital economist and focus on the specific elements of compensation considered by the interim rules and the types of families against or for whom your rules are biased and why:

1. Earnings

A Columbia University economist initially developed the specification of the human capital earnings function in 1956. Finis Welch, the former editor of the American Economic Review, has referred to this specification as "probably the most widely accepted empirical specification in economics". The alternative approach proposed by your advisor assumes an annual increase of 6.6 percent to age 30, an annual increase of 5.1 percent to age 50 and a 4.2 percent annual increase beyond age 50. Using the human capital earnings function, the annual increases in earnings during the early years can exceed 10 percent, while those in the later years can approach zero. By using the ad hoc approach, proposed by your economic advisors, you are biasing the projections, undercompensating young decedents and overcompensating older ones. The point is, by ignoring basic human capital economics, you have created more complexity and have introduced a bias in your estimates.

One of the fundamental tenets of human capital economics is that the rate of increase in earnings over the life cycle depends on education. A casual review of the census data for earnings by age and education will demonstrate its impact on the shape of the life cycle earnings curve. Your approach, supported only by the self-proclaimed experts from litigation, undercompensates for higher education and overcompensates for the lack of it.

Another factor impacting life cycle earnings is occupation. Those employed in the financial services sector, for example, have much steeper life cycle earnings curves than most other professions. As is the case with analyses prepared by most litigation economists, your rules do not account for the effect of occupation on lifetime earnings.

The primary reason that not all people work on a full-time basis to a normal retirement age is that they die. This affects the average worklife of all workers. A forty-year old male has roughly an 80 percent chance of living to age 65. This has not changed much since the late 1970s. However, it may be true that using the old worklife tables understates the worklife of women. Further, the effect of education on worklife is ignored. It is well established that higher education leads to longer worklife. By ignoring this fact, your results are biased against more highly-educated decedents.

2. Income Taxes

Taxes cannot be determined without considering all household income. While it is important to identify that income that ceased with the death, it also is important to include income that does not end when estimating taxes.

3. Consumption

Although most economists involved in litigation are unaware of it, the definition used by the Bureau of Labor Statistics (BLS) in the Consumer Expenditure Survey is "household income", defined as income from "all sources". This includes spouse's income and income from dividends, rents, annuities and other sources. Consequently, if you determine your percentages for consumption and apply them only to the income of the decedent, you will understate the actual consumption. This means that decedents with investment income or whose spouses were employed will be overcompensated by the rules.

You also should be aware that consumption expenditures as a percentages of income declines as household income rises. By ignoring this, you overstate consumption and, hence, undercompensate those with above-average incomes.

Finally, the interim rules appear to suggest that the lost earnings are first brought to a present value, and that then a consumption percentage is applied thereto. This would appear to make it impossible to use a different consumption percentage for each year. However, the rules suggest that different percentages will be used for families, depending on the number of children in the home under the age of 18. This would only be possible if consumption is determined on a yearly basis, with the net remainder discounted to present value.

4. Discounting

Long-term tax-exempt municipal bond rates vary from 2.2 percent to 5.0 percent. Taxable Treasury rates vary from 2.9 percent to 4.9 percent (5.4 percent for 30-year bonds). The rate of return depends upon the duration of the payment stream desired. Consequently, the appropriate discount rate will vary with worklife. Your discount rate exceeds even the current long-term rate of taxable Treasury bonds. This makes no sense. Your discount rate is too high, even if you are discounting a long-term taxable net income stream. In fact, you are discounting an after-tax stream. Therefore, your discount rate is too high for two reasons, and this means that families are being undercompensated with a lump sum award. Moreover, by applying a single discount rate, you exacerbate the overdiscounting, especially for older decedents.

Finally, a key element in the selection of an investment portfolio to create a year-to-year stream of income is uncertainty. A basic principle from financial economics is that the greater the volatility of an income stream one expects from an investment, the greater is the required rate of return to reward those accepting that fluctuation. In contrast, if you are "discounting" wages, which generally move in a steady and predictable fashion, discount rates associated with risk-free investments, such as Treasury bonds, are more appropriate. However, if the stream being discounted fluctuates widely, such as would be expected if bonuses, commissions or stock options are an important part of one's annual earnings, overcompensation would occur if the lost stream were replaced with risk-free treasury bonds. For such individuals, a higher, risk-adjusted "discount rate" would be appropriate.

5. Periodic Payments

You make the receipt of periodic payments discretionary on the part of the claimants. One can argue that they should not have the option of receiving a lump sum payment. Their financial loss is not a lump sum loss. They did not lose the present value of the future earnings of the decedent. They lost the future stream of those earnings. By paying out an income stream over time, rather than making a lump sum payment, the cost to taxpayers could be reduced by from fifteen to twenty-five percent. In this manner, you could pay more for non-economic damages, you could increase your "cap" above $231,000, or you could let the families keep their life insurance. It makes little sense to make an effort to hold down the cost to the taxpayer, yet omit such an obvious source of savings. You should not force taxpayers to pay more by converting the future lost income stream to a lump sum. If the recipient takes a lump sum, they are faced with the problem of recreating the lost earnings stream by making their own investments as well as the temptation to focus on short term ends to which a large lump sum could be devoted. A periodic payment fully compensates claimants, eliminates the need for reinvestment, protects at least some against the temptation to squander what might appear to be a large sum of money, and minimizes the compensation costs to taxpayers.

You have been criticized for paying more to some than to others. This becomes less obvious and more understandable if you are simply replacing the year-to-year loss on a periodic payment basis. If the focus is on "the number", unsupported comparisons will catch the headlines.

6. Active Versus Passive Income

Passive income does not cease with death. However, it does impact taxes and consumption. Be careful to distinguish between these two types of income sources. In instances of very high passive income and very low active income, death can lead to zero financial loss to a family.

7. Cap on Earnings

Perhaps the least defensible element of your rules is the cap on earnings. While it is understandable that you would want to hold down the cost of compensation, this is not the way to do it. Clearly, incomes above that cap are going to be associated with older workers. Your cap, if you insist on retaining it at the 98th percentile, should be determined by age. That is, use the 98th percentile for workers the same age as the decedent, with perhaps the same education and occupation to avoid any bias.

8. Life Insurance

It is understandable, because of the statute, that you believe you must subtract life insurance, regardless of constitutionality issues that may be raised in the future. But consider that you have the authority to "add" an amount to the total "award" that is equivalent to the amount of life insurance. Then subtract the life insurance, thereby avoiding the penalty for those families receiving life insurance benefits. Furthermore, the purpose of life insurance is not limited to replacing the economic loss. (In fact, the life industry does not even sell life insurance to cover financial loss. Unfortunately, the industry bases its coverage recommendations on their clients' "needs" rather than expected lifetime earnings. Needs-based selling focuses on taking care of current expenses rather than replacing the loss of an expected earnings stream. Consequently, there is little relationship between the amount of life insurance the industry recommends for a family and the loss associated with the death of the insured.) The fund's rules make it appear that life insurance is a luxury rather than a way to protect against the loss of a lifetime earnings stream..

The Life and Health Insurance Foundation for Education (LIFE) has a web site,
http://www.life-line.org/life/life_value.html , that provides an application of the approach from human capital economics to estimate the economic loss to a family caused by death. While not including education or specific occupations, it applies the economics of human capital, such as life-cycle earnings as specified by the standard human capital earnings function from economics and calculates consumption in a manner consistent with the BLS definitions. By entering information on specific individuals, you can see how the interim compensation rules bias the results.

The bottom line is that (1) you are spending more taxpayer money than necessary by converting losses to a lump sum rather than providing periodic payments, and (2) your rules tend to undercompensate families of decedents who were highly educated, had higher earnings and were younger. They tend to overcompensate those who had significant investment income or other passive income and those whose spouses worked. The ironic fact is that you gain no simplicity, as the correct techniques are no more difficult to apply and would eliminate the above-mentioned biases.

I am confident that the issues discussed herein are now well understood by lawyers in the Department of Justice's Torts Division. Suffice it to say that, by endorsing the self-proclaimed "economics" of those who do not recognize or make use of accepted economic analysis of human capital loss, you are failing to adequately compensate thousands of claimants. Moreover, you are endorsing those who turn their backs on fifty years of economic research because it is expedient to do so. The fact that they can make a living testifying to non-economists does not make their ad hoc approaches meaningful for your task. This can have an impact on cases well after your task has been completed. It is easy to predict the type of cases that will not be filed with the fund and, therefore, will be moved to file against the airlines.

When you take science out of a process you create bias. The procedures embodied in the interim rules were clearly designed without the benefit of the appropriate economic approach. Do not defend the interim rules by pointing to their supposed simplicity. The correct approach is not more complicated, and it will produce results that are both consistent and that can be defended on solid economic grounds. I wish you the very best in the task that lies ahead.

Individual Comment

2002-01-18