Semi-Daily Journal

Why am I not surprised. Doesn't Boskin get a black eye for having his analysis debunked? I have seen some recent numbers on 401K and they are not pretty.

According to a Hewitt Associates survey of 1.5 million plan participants, only about 17 percent of workers made any investment trade in their 401(k) plan during 2002. That was the third year in a row in which most workers didn't add any money at all to their 401(k) retirement plans. More disturbing is that workers are abandoning their plans. The proportion that participated in a 401(k) plan fell to 70 percent last year, from 80 percent in 1999, Spectrem Group reports.

Other 401(k)'ers, however, aren't heeding the advice of experts who warn against dangerously overloading retirement plans with company stock. Huge losses suffered by workers at companies in bankruptcy protection, such as Enron. should not be forgotten.

Nearly 30 percent of plan assets are still invested in employer stock, according to an analysis of 329 plans in June by the Institute of Management and Administration. At 58 of the companies, workers have 50 percent or more of their plan assets invested in company stock.

I would not be surprised if there is less money in these plans than predicted. Tired of seeing their accounts go down, people are pulling out and investing money elsewhere, such as buying bigger houses, etc. Many 401s limit the options available. Bond funds are overpriced. Stocks are way overpriced. Money Market returns are next to worthless. Many investors would be better off buying I-bonds (which are not usually available within a 401k) than buying stocks at overinflated prices. If we are in a secular bear market comparable to the 1970s, then these accounts will not gain much in value.

Looking at the national debt numbers, the US was in deficit (including money owed to SS) over $420 Billion from 2001 to 2002. We are going to top that this year. How can 401s possibly pay enough future taxes if less than 1 in 5 participate?

When I read Boskin's paper the one simple thing that jumped out at me was that he assumes a 27.7% average tax rate on these savings, when the actual average tax rate is 16% and he notes that the average tax rate for retirees is less than the average rate overall.

Being that tax brackets are indexed for inflation the 27.7% tax rate he assumes indicates average taxable retirement income for people owning such accounts of well over $200,000 per year in today's money -- if we're all that rich in the future we really won't have any problems.

He figured his average rate using the self-reported data in the Fed's Survey of Consumer Finances for persons who report having retirement accounts, while noting there are a lot of problems with self-reported data.

Now I didn't follow up on this in any depth, so I don't know and could be wrong, but I have a sneaking suspicion that people in the survey with traditional defined-benefit pension plans provided by employers significantly under-reported them. After all, they can't know how much their plan accounts are currently worth years before retirement (unlike with defined contribution plans) and don't think that the plan assets are "theirs" in the sense that IRA or 401(k) assets are -- the assets belong to the GM pension plan, or whomever.

These people tend to be lower-tax bracket than people with large amounts in "owned" plans, so their underreporting would bias the survey's average tax bracket rate for people with retirement savings upward. (I.e.: If you have $500,000 in an IRA you know it and report in the survey that you are in the 30% tax bracket, but if you have $150,000 in the GM pension plan you don't know it, don't report it in the survey, and don't report that you are in the 15% tax bracket, so your lower tax rate is eliminated from the averaging computation.)

So it seems possible that Boskin took correct numbers for total accumulations in retirement plans from industry data and such, while figuring the average tax rate from self-reported survey numbers that are biased high, resulting in a significantly too high estimate of the tax rate to be applied to collect future tax revenue from the accounts.

Well, I dunno, I could easily be wrong about this suspicion. But whatever is going on, assuming a 27.7% average tax rate that is so much higher than the actual 16% average rate, when brackets are inflation-adjusted and with no other explanation, raises questions to me.

My first reaction to Boskin's paper was that his claim of finding something new could not be because others have discussed this issue to some degree. My second reaction was that any generational accounting forecast that forgot this simple point could not have been done by a truly smart person. And it would seem the smart generational accounting types did not make such an obvious error. My third reaction was wondering (sometimes outloud) whether Boskin overstated his case with bad finance and economics. I'll have to read the most recent paper more carefully along with Boskin's as the authors of the recent paper typically do first rate finance and economics.

"How can 401s possibly pay enough future taxes if less than 1 in 5 participate?"

Boskin isn't considering just 401(k)s but government pensions, military pensions, life insurance annuities, traditional defined benefit pensions, IRAs, etc. -- everything that's tax deferred. IIRC 80% of workers over age 40 have a least one of these.

They sure don't have an average 27.7% tax rate though. Looking quickly at his paper again, the survey data he uses to compute the average tax rate for persons with such savings only covers about 1/3rd of the total, excluding the likes of government pensions and traditional defined-benefit pensions, and is seemingly quite biased towards high-value accounts (and hence top-tax-rate account owners.) I still don't see any sign that his tax rate isn't 30% too high.

"Boskin's projections of revenue from tax-deferred accounts have only a very modest effect on the long-term fiscal outlook because almost all of the relevant revenue is already incorporated into the revenue projections", and

"... any generational accounting forecast that forgot this simple point could not have been done by a truly smart person. And it would seem the smart generational accounting types did not make such an obvious error."

Well, one would think so, but Boskin reports otherwise and that is what makes his paper so startling. To quote him...

"...it is surprising, perhaps disconcerting, that virtually no information on deferred taxes beyond short-run historical flows is currently available in any form, anywhere. Not in the Fed's Flow of Funds sectoral balance sheets. Not in Treasury Financial Statements of the U.S. Government. Not in OMB's Analytical Perspectives on the Budget. Not in academic research."

So that seems a pretty clear point of difference between him and Auerbach/Gale/Orszag and their statement above.

Either Boskin is flat wrong about the information not being available "in any form, anywhere" before this, which would appear quite a howler, or they have to say that even though the information wasn't available in any form, anywhere, it was still somehow "almost all" automatically incorporated into revenue projections anyhow.

It will be interesting to see how this turns out.

BTW the Boskin paper is at http://emlab.berkeley.edu/users/burch/e231_sp03/Boskin.pdf

As an interesting side light, William Gale was Mike Boskin's research assistant when he was a graduate student at Stanford.

I'm surprised to hear that I-bonds aren't available to most 401(k) plans-- I note, e.g., they're available in my plan through a Vanguard fund.

Suggest taking a look at Stephen Roach of Morgan Stanley. We have a serious serious serious savings problem! We are likely to have zero to negative national savings, public and private combined, in the coming year.

Jim Glass writes "Boskin isn't considering just 401(k)s but government pensions, military pensions, life insurance annuities, traditional defined benefit pensions, IRAs, etc. -- everything that's tax deferred. IIRC 80% of workers over age 40 have a least one of these."

I haven't read Boskin, nor am I likely to. He strikes me on his past record as being a stupid right-wing flack, but without Arthur Laffer's great artistic ability or Jude Wanniski's deep understanding of economics.

Still, I think Jim's squib as it stands shows two errors in Boskin's general thesis. First he is looking at huge capital accumulations which have been assembled as deductions from taxable income, and which have accumulated without ongoing income or capital gains taxes, and he is taxing those accumulations as if, to quote the myth, "the baby boomers are liquidating their holdings."

This does not, and will not happen. The poorest few deciles of retirees may draw down their savings, reverse mortgage their homes, sell their trailers, and so forth. There does not exist, and never has existed, a retired cohort which overall drew down its capital in retirement. A cohort's, the people of a given birth year's, marginal propensity to invest increases monotonically with age. Many retirees may be poor, and some may get poorer, but retirees as a whole live on their income and get richer as they get older.

There is no draw-down of previously accumulated tax free income which is subject to recapture of previous taxes. Retirees will continue to pay income tax, but generally the income they take from their investments will be lower than that which they had as workers. Boskin's supply of funds which suddenly become taxable is simply non-existent.

The only way to claw back the taxes "deferred" over the working lives of retirees is through death duties on accumulated capital, once they're gone and don't need the income. Inconveniently for his thesis, Boskin's buddy Bush is cutting off this possibility.

Second, Jim's apparent suggestion that all other forms of retirement income share some sort of taxability with the capital amounts inside 401k's, is incorrect. The accumulated reserves which make it possible for an insurance company to pay off an annuity or a pension do not become taxable in any way simply because the owner of the pension or the annuity starts collecting. They remain reserves and investments of the insurance company, and are not subject to tax.

Military pensions are not vested, and the amount of the pension -- always lower than the working income it replaces -- is only subject to normal income taxes.

In other words, there is no new and undiscovered source of tax hiding in that wider set of pensions.

Sorry, folks: it's all a dream.

" There is no draw-down of previously accumulated tax free income which is subject to recapture of previous taxes."

Didn't you used to play a CPA on usenet, David? There are stiff penalties for people who do not take the MINIMUM REQUIRED withdrawals from their tax deferred investment vehicles (I believe you divide the total amount of the vehicle by your remaining life expectancy to calculate that minimum)

" The only way to claw back the taxes "deferred" over the working lives of retirees is through death duties on accumulated capital, once they're gone and don't need the income. Inconveniently for his thesis, Boskin's buddy Bush is cutting off this possibility."

Sory, wrong again, and as usual. The beneficiary is under the same requirement to withdraw a minimum amount every year.