Sunday, January 13, 2019

Why You Can't Retire

I recently conversed informally with a financial representative of TIAA-CREF.  He is in his thirties, and he expressed concern that the Social Security retirement age would be raised to 80 or greater before he retires.  I agreed that the age would be raised, but I suggested it might not be as bad as 80--perhaps the ultimate age would be 69 or 70.  Unfortunately, the problem will become worse if it is not addressed currently, and the media and the public have not been overly interested.  Hence, the broad explanation as to why you can't retire is inattention at both the personal and public levels.  Public policy that has hammered down savings rates and interest rates and has helped wealthy investors with strong risk appetites hasn't helped.

Because of monetary stimulus, low interest rates, inflation, and money illusion, along with escalating health care costs, many Americans have earned low wages.  Money illusion and health care cost inflation have caused productivity growth to exceed wage growth.  Also because of inflation, many Americans have avoided saving.  Lower-income Americans are not used to investing in the stock market and so have not benefited from the Federal Reserve Bank's inflationary reallocation of wealth from wages and savings to capital and stock-and-bond investments.

The Fed's pumping up of the stock market for the past three-and-a-half decades has exacerbated a natural tendency for the market to rise when a disproportionate share of the population (the baby boomers) is young. (In 2017 the Census Bureau found that the average age had increased since 2000 from 35 to 37.) Although the market averages are still high, in 2019 the majority of stocks have moderate price-earnings ratios. That may be consistent with an underlying downtrend in valuations as the average age increases. What's propping up the S&P 500 are a few bubble stocks like Amazon.com. A declining stock market in the boomers' later years might make retirement difficult.

High tax, health cost, and rent burdens haven't helped.  The Peter G. Peterson Foundation finds that median taxpayers pay about 14 percent of their income in federal taxes. The Tax Foundation says that median taxpayers pay about 9.9 percent in state and local taxes. Hence, about 25% of income is paid in taxes. That is low in comparison with other OECD countries, but the majority of Americans pay for their own health insurance, and the $18,000 cost  for a family is about thirty percent of the $61,000+ median household income as of 2017.  Of that, about $6,000 is paid out-of-pocket. The incidence of the $18,000 cost is likely on employees as is the 6.2 percent employer Social Security tax (see the 1971 American Economic Review article by John Brittain).   If one adds the 25 percent of income paid in taxes to the 10 percent of income paid for out-of-pocket health insurance costs plus the 6.2 percent employer Social Security tax, the tax burden to employees is above 38%. (The base needs to be grossed up for the imputed employer Social Security tax.) If you include the $18,000 family health insurance cost, the tax-and-insurance burden is close  to 48 percent. (Again, the based needs to be grossed up for the imputed non-out-of-pocket health insurance costs and the employer Social Security.) That's a little lower than the highest-tax country, Belgium (54 percent), but about equal to the second-highest-taxed country,  Germany (48 percent).  Hence, contemplating the incidence of the employer Social Security tax and family health insurance cost places the US in the top three  of OECD countries with respect to tax burden.

Given that Federal Reserve Bank policy has been for the Fed to pump up asset values, rents have  escalated. According to the Census Bureau, rents in 2000 dollars rose from $284 in 1940 to $602 in 2000.  According to Household Income, rents increased by 12 percent from 2000 to 2010, but the median income fell by 7 percent. In other words, the real cost of rents has more than doubled while real wages have remained stagnant.

Federal policy has been to burden US employees with high taxes (when supplemented by health insurance and employer part of Social Security costs) and to subsidize landlords with escalating property values that boost rents by several percent each year.  How many wage earners can save given that constellation of policies?

As well, Social Security is insecure. According to Steven C. Goss of the Social Security Administration, taxes will cover only 75 percent of benefits by 2035.  (Note that Goss does not use the word contributions, which is appropriate to a pension, or the word premiums, which is appropriate to insurance.)  As well, even if the problem in 2035 is solved, there is an additional problem looming in 2080, when infants born in the next few years retire.  There is no talk about solving the later problem.

With respect to the 2035 problem, Goss says that if changes are made immediately, there will need to be a benefit reduction of 13 percent or an increase in the FICA tax rate of two percent (from 12.4 percent to 14.4 percent) to cover the shortfall.

The average American retires at 62, the first year of eligibility for Social Security retirement benefits. According to the Social Security life expectancy table, life expectancy at 62 is 20 years; at 66, the current normal retirement age, life expectancy is 17.1 years;  at 70, life expectancy is 14.3 years. To reduce benefits by 14.7 percent for those first eligible to retire, the minimum retirement age of 62 would need to be raised to 66.  To reduce benefits by 12.3 percent for those who retire at the normal retirement age (currently 66 but scheduled to increase to 67), the normal retirement age would need to be raised to 69 or 70.   The precise amounts would be adjusted by actuarial calculations, but a three- or four-year increase in the retirement age would solve the issue.  

The Social Security benefit structure favors low-wage workers because the salary bands have higher percentages at lower bands than at higher bands. That is,  as of 2019, the benefit based on the highest 35 years of indexed monthly earnings is calculated as  90% of the first $926; 32% from $926 to $5,583; 15% of the monthly earnings greater than $5,583.

Someone having 35-year indexed earnings of $2,083 a month will get  $1203 a month in Social Security ($14,400 a year) while someone having a 35-year indexed earnings of $8,333 will get $2,569 (30,828 a year).   The $1,203 is 58% of the earnings of the person who makes $25,000 per year; the 2,569 is only 31% of the earnings of the person who makes $100,000 per year.

Hence, the benefit structure is subsidized in favor of low earners, but it is meager across all income levels. Moreover, the Brookings Institution argues that the welfare subsidy to lower-income workers should be increased because they suffer from lower life expectancy and decreasing lifetime earnings. Brookings advocates increasing the welfare plan essence of Social Security by reducing the rate for the top band.  

The Heritage Foundation also suggests supplementing Social Security's  welfare component. The Heritage Foundation suggests reducing benefits and contributions for those whose retirement incomes are above $55,000 and eliminating it for retirees whose earnings are over $110,000. For those born after 1985 Social Security would be turned into a flat payment.

Income during retirement is primarily from six sources: stock dividends, savings accounts, bond interest, 401(k) plan distributions, defined benefit plan distributions, and IRA distributions. However, retirees can reallocate wealth into investments, such as growth stocks, that chiefly pay capital gains and so sidestep the maximums.Thus, rich retirees could still collect Social Security if they prefer.

It might make more sense to simply face its welfare essence and to eliminate Social Security for those who have above a minimal level of income. Perhaps a mandatory public-private target benefit plan approach could be used so that benefits are funded though a combination of sources depending on income level. A certain benefit could be targeted or defined, such as 80 percent of income at age 70,  based on long-term assumptions, but the actual benefit would depend on the amount in the fund.  For those earning below the average wage each year, a small out-of-pocket contribution, say one percent, could be supplemented with a welfare subsidy.

For instance, for those earning half the average wage, the government might make all of the contribution; for those earning three-quarters of the average wage, the government might make half of the contribution; for those earning the average wage and above, the taxpayer might make the full contribution. An contribution as percentage of pay actuarially expected to achieve 80 pension income replacement could be invested in ways that the taxpayer chooses. Perhaps the investment choices could be vetted to ensure legitimacy, although such a review likely would have passed Enron and Worldcom with flying colors.  Perhaps investors might be permitted to change investment direction only once every five years and not be permitted to make interfund transfers. There's a bit of nanny state in this, but eliminating centralized control through privatization will provide better choice.

The usefulness of a Social Security trust fund is doubtful. If you had invested $4,000 at six percent interest from 1979 until today, you would have $619,000.  That would buy an annuity of over $40,000 at today's rates. In other words, it's far from clear that Social Security is a good deal. Taking investment policy out of politicians' hands would help the economy and help investors. 

Perhaps retirement policy might be downloaded to each state, allowing the states to determine preferences that differ locally.  New York or California, for instance, would probably prefer a higher welfare component while Idaho, Utah, or Georgia would probably prefer a lower welfare component.  Rather than one region's imposing its preferences on another, a decentralized approach would allow expression of local preferences.