COAUTHORED WITH DOUG WISSOKER
A recent paper by Bayer, Ferreira, and Ross on mortgage delinquencies and foreclosures finds that people of color had greater problems once Recession hit than did many others in roughly equal circumstances, such as income and location, but with different racial backgrounds. We believe this is a useful, though not surprising, finding in ongoing studies of the impact of the Recession on different types of households. Yet we worry about how its results get extrapolated into policy recommendations.
The paper concludes that their research “raises concerns about homeownership as a vehicle for reducing racial wealth disparities”. We believe that one needs to be very careful in extrapolating lessons from the market of the mid-2000s to any market and to policies that would apply over time. Paying off mortgages is the primary means by which the majority of households, particularly low and moderate-income households, save over time. Discouraging such saving could easily add to already unequal distribution of wealth in society.
First, a quick summary of the findings. Combining several sources of data to look at racial differences in delinquent payments and foreclosures for mortgages for purchases and refinances originated between 2004 and 2008, the authors find that black and Hispanic borrowers had substantially higher delinquency and foreclosure rates than whites and Asians, even controlling for differences in circumstances such as the borrower’s credit score, the size of the interest rate spread of the loan, and the identity of the lender. In addition, the authors conclude that the racial gap in delinquent payments and foreclosures peaked for loans originating in 2006. From this, they conclude that people of color entering the market at the peak of the housing boom were particularly vulnerable to adverse economic conditions.
The authors attribute the racial difference found for blacks and Hispanics, even after trying to control for income or other differences, to items they couldn’t measure, including lower wealth and an accompanying lack of a financial cushion. This seems crucial to us and is also consistent with studies that income an incomplete predictor of upward or downward mobility. Work from the Urban Institute (here) shows that wealth differentials by race are much greater than income differentials. These differentials can play out in multiple ways across generations. For instance, wealthier families provide more inheritances and intergenerational transfers that support homebuying and downpayment levels that reduce foreclosure risk.
However, the authors’ concern about homeownership as a vehicle for reducing racial wealth disparities does not follow logically. Evidence here is at best circumstantial. Among other sources of disparate outcomes, consumer groups would point out that these types of findings more than anything highlight the disparate impact of abusive lending at the height of the housing boom.
Portfolio theory requires looking across different types of assets and debts, along with their associated expected returns and risks. Homeownership has risks, but so does renting. In fact, rental rates at times rise faster than the costs of homeownership, and in many parts of the country it has become cheaper to own than rent for those likely to be in a home long enough that transactions costs do not eat away at the ownership returns. Similarly, a household often must choose among debt instruments. Mortgages tend to have lower interest charges than most other forms of debt.
Most vehicles for getting a decent return on investment involve some risk. Saving accounts now paying negative, after-inflation, returns only prove the point in spades. If saving were proportionate to income, for instance, but lower-income individuals invest only in low or negative return assets, then wealth inequality necessarily would grow to be much greater than implied by levels of saving, potentially compounding adverse outcomes over time. Conversely, without discounting lessons from the Great Recession, low-cost, well-structured mortgages continue to be supported by the government (whether through FHA or the GSEs) partly for the very purpose of diversifying risk and effectively spreading wealth ownership.
This study is based on patterns of delinquency and foreclosure rates observed during a limited time period with unusually high foreclosure rates. But, wealth accumulation occurs over a very long time. Thus, even on this paper’s own terms, it’s not clear that reduced rates of homeownership would make low-income households or people of color better off over extended periods. We have found that most homeowners buying a decade or so before the Great Recession came through the longer period in good shape. Our own work also tends to show that black homeownership rates, even after controlling for income, are disproportionately low in both good and bad markets, raising serious questions about whether they are missing out on opportunities available to others.
Regardless of the effect on the difference in wealth disparity by race, homeownership is an effective way for many, though certainly not all, low- and moderate-income households to save. Equity in a home is the primary asset owned by low- and middle-income households, including blacks and Hispanics, by the time of retirement. Paying off a mortgage is the primary mechanism by which these households save, with all the virtues of a more automatic and regular saving vehicle. Reductions in the already low homeownership of people of color would almost certainly exacerbate over time the unequal distribution of wealth.
Knowing how many of us economists toil away in obscurity on most research, I’m always intrigued by what catches the press’s and public’s attention. Take, for example, the significant attention paid to a 2010 study by Harvard economists Carmen Reinhart and Kenneth Rogoff that concluded that countries with debt levels above 90 percent of GDP began showing slower rates of growth. When Thomas Herndon, Michael Ash and Robert Pollin, scholars at the University of Massachusetts at Amherst, recently had trouble replicating Reinhart and Rogoff’s results, the debate played out in national news outlet.
Unfortunately, this discussion quickly devolved from substance to politics to arguments ad hominem. Without getting into the extent to which I or others can validate Reinhart and Rogoff’s (R&R’s) original findings, I offer six cautions for anyone witnessing this or a similar statistical debate with significant policy implications: (1) statistics should never be interpreted as showing more than simple but potentially useful correlations; (2) healthy skepticism is required for all social science research, which seldom gets replicated for validity; (3) all empirical economic work is based on history that will not repeat in the exact same form; (4) research can certainly contradict conventional wisdom but not reason; (5) arguments ad hominem, particularly by those with their own agendas, are unhelpful; and (6) be careful with labels and straw men.
- Correlation versus causation. It’s long been stressed that statistical tests never prove causation, not simply because they can’t but because researchers make many choices and assumptions, often of statistical convenience. This doesn’t mean statistics are useless. Just accept that any result merely shows that A and B seem to occur together even after trying to account for other influences under a huge range of never-fully-tested assumptions.
- Skepticism. A growing body of “research on research” shows that few social science experiments, and even many medical studies, are replicated. Also, positive results get published; negative results usually do not. R&R’s study was replicated mainly because it got an unusual amount of attention.
- History. The past never repeats itself exactly—or, as Heraclitus warns, “You could not step twice into the same river.” That historical interpretations are contained and sometimes couched in data analyses doesn’t mitigate this well-known caution.
- Reason versus conventional wisdom. The R&R debate mainly revolves around two reasonable notions. One is the simple arithmetic conclusion that debt can’t rise forever relative to national income, along with the related economic conclusion that higher levels of debt can and have been shown in many places to have consequences for investment, interest rates, ability to borrow, and how government revenues are spent. The other is that institutions, times, places, and circumstances matter greatly, and they affect how one should interpret past data, such as those presented by both R&R and their critics.
- Ad hominem arguments. R&R published some of their work with the Peterson Institute for International Economics, so some of those who attacked R&R may have considered the authors guilty by association because of Peter George Peterson’s concern about deficits and his contributions to that Institute. Peterson’s own “guilt” on budget issues seems to be that he became rich on Wall Street, although he favors higher taxes on what he calls “fat cats” like himself. Still, while many of those who engage in these attacks themselves fail to represent their own or their institution’s sources of funding, let’s be honest. Much social science research is funded in ways that doesn’t necessarily bias how the research is done, but rather what is researched in the first place. So, unless one fully engages and thoroughly analyzes every study, even the careful academic reader must often try to determine trustworthiness in other ways, such as whether the researchers report results only consistent with some special interest or political party.
- Labels and “straw men.” While we all use labels and straw men at times to set up the stories we tell, they at best simplify greatly. In this case, R&R are identified as advocates of “austerity” and their opponents as “Keynesians” advocating stimulus, both of which are nothing more than labels. Most budget analysts I know worry about deficits but vary widely in whether they would engage in more short-term stimulus or in how strongly they believe that a path toward long-term balance even requires austerity. For instance, is reducing some rate of growth of spending, no matter what its level, austerity? Is the Congressional Budget Office an advocate of austerity or Keynesian when it asserts that sequestration hurts the economy in the short run, but has long-run benefits relative to doing nothing about deficits?
The bottom line: use extreme caution no matter which economist you read or believe.
Full disclosure: I have spent most of my career at the Urban Institute or the Treasury Department, brief periods each at the Brookings Institution, American Enterprise Institute, and the Peter G. Peterson Foundation (which differs from the Peterson Institute for International Economics). I also serve or have served on many advisory groups and boards for such organizations as the Committee for a Responsible Federal Budget and the Comptroller General of the United States. As a consequence, je m’accuse of being among the many economists limited more than I would like by what research is supported by those institutions or their funders.
Arithmetic tells us we must either decrease the growth of Social Security spending or increase taxes as a share of gross domestic product.
But we should do it with an eye on fairness, growth and efficiency. We’re all in this together, so higher-income families must give up something to deal both with Social Security shortfalls and those in the budget more generally. A modest increase in the wage base for Social Security has some justification since that base has eroded in recent years. But if extended too far, it exacerbates the squeeze on other government programs. How? On the tax side, it tends to preempt other tax increases for non-Social Security purposes. On the benefit side, it attempts to maintain a growth rate of Social Security and other elderly programs that absorb more than all of the scheduled growth in government spending for decades to come, thus continuing a downward spiral in the share of the overall budget devoted to children, education and investment more generally.
Under current Social Security formulas, ending the cap on income would mean that some fairly wealthy individuals would get benefits in excess of $1 million. Though no one thinks that that makes sense as a benefit schedule, capping benefits goes against the Social Security tradition of being paid back for additional contributions. On the technical front, an unlimited Social Security tax would also encourage individuals to reclassify labor income as capital income not subject to Social Security tax. This would be a special problem for the self-employed and owners of partnerships, since Social Security now taxes both their capital and labor income as labor income.
Finally, the Social Security Administration’s Office of the Actuary found that even with a cap on benefits, the wage base expansion would still leave the program running future deficits. We shouldn’t pretend that it does otherwise.
This column was reposted from New York Time’s Room for Debate.
While the increase in dementia among the elderly and the president’s proposal to change the index used to provide cost-of-living adjustments (or COLAs) to Social Security recipients have both received prominent headlines recently, the discussions have largely been independent of one another. Yet any principled attempt to reform our elderly programs, including Social Security, Medicare, and Medicaid long-term care, should consider them together.
A well-designed reform of elderly programs could and should accommodate some of the cost problems associated with dementia by back-loading a larger share of benefits in Social Security to older ages when these and other needs of old age increase. COLA adjustments, whatever their other merits, front-load the system by cutting back on benefits for the oldest the most and those in late middle age or their 60s hardly at all. That the president and Congress have limited ability to engage in these types of discussions and tackle multiple goals at the same time is yet one more example of how our political processes increasingly block us from fixing what ails us.
In a well-cited RAND study, Michael Hurd and his coauthors estimate that dementia-related care purchased in the marketplace will cost somewhere close to $0.25 trillion in 2040 (in 2010 dollars). That sounds like and is a lot of money, but Social Security and Medicare are expected to rise to cost over $3.5 trillion in that same year. Although I am greatly simplifying by ignoring such factors as how much of the $0.25 trillion would be covered by individuals and not the government or the effect of entitlement reform on costs, the raw comparison speaks for itself.
Simply put, some of the private and public budget problems associated with dementia, Alzheimer’s, and other growing problems for the older among the elderly could be addressed by providing higher cash benefits in older ages. Whatever the aggregate size of Social Security in general, one could pay for this reform by cutting back on benefits in younger ages of Social Security “old age insurance” receipt. This would not solve all the associated problems of dementia, but it would be a simple, effective, easy-to-administer step in the right direction. And, by concentrating benefits more in older ages, it would encourage working longer at a time when employment rates for the population as a whole are scheduled to decline.
But this is not the discussion we’re having. Instead, the president and many budget reformers put forward a proposal to adapt what many believe is a better measure of cost-of-living or price changes and apply it to almost all government programs, including Social Security. As a technical matter, a COLA adjustment doesn’t affect the growth in initial Social Security benefits for those who retire, only the inflation adjustment they get after they retire. At that point, they get a small annual cut—e.g. 3/10 of 1% the first year, 6/10 of 1% the second year, and so forth—that compounds every year in retirement, so that by the time beneficiaries are in their late 80s or 90s, some 25 or 30 years of lower COLAs add up to a cut in benefits of as much as 10 percent.
Social Security has never adjusted upward the earliest retirement age for increases in life expectancy. Instead, it reduced the earliest age from 65 to 62 in 1959 and 1962. As a consequence, the share of benefits going to those with 15 or more remaining years of expected receipt has risen dramatically over time, and the share to those with, say, less than 10 years of remaining life expectancy has declined. The COLA proposal, even with some very old age adjustments suggested by the president, would add to this long-term trend of making the program ever less available in relative terms for those in truly old age.
This is not to say that the COLA proposal should not be adopted. Who can oppose trying to measure something better? But attempts to fix systems like Social Security and other elderly programs one parameter or adjustment at a time cannot easily meet multiple worthwhile objectives. Similarly, efforts to back-load the system to meet the needs of true old age, as suggested here, should be coordinated with further adjustments—say, in minimum benefits—to avoid discriminating against those with shorter life expectancies.
With or without a better COLA, therefore, reform of Social Security and other elderly programs requires a more comprehensive approach if we are to meet the needs of old age as they evolve over time. Shouldn’t dementia be a higher priority than early retirement? If we’re going to spend $3 trillion or more annually on Social Security and Medicare by 2040, do we really think that the allocation of those funds be determined by formulas set in years like 1935 or 1965 or 1977, when much of the current system was cobbled together?
My colleagues and I recently published research showing that younger age groups are falling behind their parents in wealth accumulation and explaining the story behind our numbers. Some have raised questions about how we use our data, and I want to take some time to further explain our research.
Our study shows that the average wealth, or net worth, of these younger age groups has fallen fairly dramatically relative to older age groups. In response, some have said that median wealth is more important than average wealth. In fact, both are important. Average wealth tells us how a group is prospering as a whole relative to other groups; median wealth tells us how some “typical” person might be doing. One complication with focusing on median wealth is that it doesn’t show where all the remaining wealth goes. In a similar vein, if you were studying small business ownership by age or race, the median value might be zero for all groups. The average values would be greater than zero and thus would allow comparisons by groups.
Consider the median household age 56–64 in 2010. True, it is only slightly richer than the median household of a similar age in 1983 ($179,400 versus $143,150). Still, the median household age 29–37 in 1983 had $46,234 in wealth, but the median household in that age group in 2010 had only $15,900, less than half compared to their parents.
Median and average net worth by age is reported here. Come to your own conclusion.
Another footnote: Our study did not look at the decline in defined benefit wealth. However, the availability of such wealth has declined more for younger than older groups. Moreover, the valuation of defined benefits and annuities goes up for those who have them when interest rates go down. Older individuals with more defined benefit wealth technically saw the value of wealth go up after the Great Recession.
You can slice and dice these data in many ways, but the empirical data speak for themselves: younger age groups have fallen behind in relative terms. All sorts of factors are involved: the Great Recession and its impact on housing, student debt, wages, and so forth. Each is worthy of our attention.
The young have been faring poorly in the job market for some time now, a condition only exacerbated by the Great Recession. Now comes disturbing news that they are also falling behind in their share of society’s wealth and their rate of wealth accumulation.
Signe Mary McKernan, Caroline Ratcliffe, Sisi Zhang, and I recently examined how different age groups have shared in the rising net wealth of the U.S. economy. Despite the recent recession, our economy in 2010 was about twice as rich both in terms of average incomes and net worth as it was 27 years earlier in 1983. But not everyone shared equally in that growth.
Younger generations have been particularly left behind. Roughly speaking, those under age 46 today, generally the Gen X and Gen Y cohorts, hadn’t accumulated any more wealth by the time they reached their 30s and 40s than their parents did over a quarter-century ago. By way of contrast, baby boomers and other older generations, or those over age 46, shared in the rising economy—they approximately doubled their net worth.
Older Generations Accumulate, Younger Generations Stagnate
Change in Average Net Worth by Age Group, 1983–2010
Source: Authors’ tabulations of the 1983, 1989, 1992, 1995, 1998, 2001, 2004, 2007, and 2010 Survey of Consumer Finances (SCF).
Notes: All dollar values are presented in 2010 dollars and data are weighted using SCF weights. The comparison is between people of the same age in 1983 and 2010.
Households usually add to their saving as they age, while income and wealth rise over time with economic growth. If these two patterns apply consistently and proportionately, then one might expect to see, say, a parent generation accumulate $100,000 by the time its members were in their 30s and $300,000 in their 60s, whereas their children might accumulate $200,000 by their 30s and $600,000 by their 60s.
This normal pattern no longer holds for the younger among us. However, this reversal didn’t just start with the Great Recession; it seems to have begun even before the turn of the century. The young increasingly have been left behind.
Potential causes are many. The Great Recession hit housing hard, but it particularly affected the young, who were more likely to have the largest balances on their loans and the least equity relative to their home values. If a house value fell 20 percent, a younger owner with 20 percent equity would lose 100 percent in housing net worth, whereas an older owner with the mortgage paid off would witness a drop of only 20 percent.
As for the stock market, it has provided very low returns over recent years, but those who hung on through the Great Recession had most of their net worth restored to pre-recession values. Bondholders usually came out ahead by the time the recession ended as interest rates fell and underlying bonds often increased in value. Also making out well were those with annuities from defined benefit pension plans and Social Security, whose values increase when interest rates fall (though the data noted above exclude those gains in asset values). Older generations hold a much higher percentage of their portfolios in assets that have recovered or appreciated since the Great Recession.
As I mentioned earlier, however, the tendency for lesser wealth accumulation among the younger generations has been occurring for some time, so the special hit they took in the Great Recession leaves out much of the story. Here we must search for other answers to the question of why the young have been falling behind. Likely candidates for their relatively worse status, many of which are correlated, include
- a lower rate of employment when in the workforce;
- delayed entry into the workforce and into periods of accumulating saving;
reduced relative pay, partly due to their first-time-ever lack of any higher educational achievement relative to past generations;
- their delayed family formation, usually a harbinger and motivator of thrift and homebuilding;
- lower relative minimum wages; and
- higher shares of compensation taken out to pay for Social Security and health care, with less left over to save.
When it comes to conventional wisdom and media attention to distributional issues, there’s a tendency simply to attribute any particular disparity, such as the young falling behind in wealth holdings, to the growth in wealth inequality in society. But the two need not be correlated. Disparities can grow within both younger and older generations, without the young necessarily falling behind as a group.
Whatever the causes, we should also remember that public policy now places increased burdens on the young, whether in ever-higher interest payments on federal debts they will be left or the political exemption of older generations from paying for their underfunded retirement and health benefits. At the same time, state and local governments have given education lower priority in their budgets; pension plans for government workers now grant reduced and sometimes zero net benefits to new, younger hires; and homeownership subsidies post-recession increasingly favor the haves over the more risky have-nots.
Maybe, more than just maybe, it’s time to think about investing in the young.
On the front page of the Washington Post on March 11, 2013, Michael Fletcher connects the different life expectancies of the poor and rich to the debate over whether Social Security should provide more years of retirement support as people live longer. He mistakenly leaves the impression that adjusting the retirement age for increases in life expectancy hurts the poor the most. In fact, such adjustments take more away from the rich. Let me explain how.
Suppose I designed a government redistribution policy that increases lifetime Social Security benefits by $200,000 for every couple with above-average income that lives to age 62. For every couple with below-average income that reaches age 62, my program would increase benefits by $100,000.
Does this sound like a good policy? Well, that’s exactly what Social Security has done by providing all of us with increasing years of retirement support. People retiring today get many, many more years of Social Security benefits than those retiring when the system was first created. And, the primary beneficiaries are the richer, not the poorer, among us. Throwing money off the roofs of tall buildings would be a more progressive policy, since the poor would likely end up with a more equal share.
Why, then, do some Social Security advocates oppose increasing the retirement age? Because the $100,000 in my example could mean proportionately more money for the poor. For instance, it might add one-tenth to their lifetime earnings (of, say, $25,000 a year for 40 years of work, or $1 million over a lifetime), while the $200,000 to rich individuals might add only one-fifteenth to their lifetime earnings. As it turns out, even this assumption isn’t correct, but let’s assume for the moment it is.
Why would we want to redistribute that way? Following that logic, we should have protected the jobs of all the Wall Street bankers after the recent crash because their wages represented a smaller share of their income than the wages of poorer workers providing support services. Or perhaps we should provide $5,000 of food stamps to those making more than $50,000 and $3,000 of food stamps to those making $20,000; after all, the latter would still get proportionately more.
As it turns out, however, more years of retirement benefits don’t benefit the poor proportionately more than the rich. Yes, the poor have lower life expectancies, but other elements of Social Security offset this factor. A greater share of the poor doesn’t make it to age 62, so a smaller share of them benefit from expansions in years of retirement support. More importantly, those who are poorer are more likely to receive disability payments that aren’t affected one way or the other by the retirement age; hence, again, a significantly smaller share of them benefit from more retirement years. Other regressive elements such as spousal and survivor benefits also come into play for reasons I won’t further explain here. Empirically, these various factors add up in such a way that increases in years of benefits help those who are richer and those who are poorer in ways roughly proportionate to their lifetime incomes.
Setting these disputes aside, the higher mortality rate of the poor at each age does raise many legitimate policy issues. Recipients who stopped smoking a couple of decades ago, for instance, have been rewarded with more and more years of retirement benefits. This, along with many other features of Social Security, such as the design of spousal benefits already noted, does mean that the system is a lot less progressive than most believe.
The more fundamental issue, then, is whether we should better protect those with low-to-average wages during their lives. I believe we should but through better-targeted mechanisms, such as minimum benefits, progressive adjustments to the benefit formula, wage supplements to low-wage workers, and other devices that don’t spend most of the program’s funds on ever more years of retirement for those who are richer.
Yet another reason to worry about the retirement age is that the failure to adjust over time—a couple retiring today at 62 can now expect about 27 years of benefits—has meant larger shares of payments go to those closer to middle age, in terms of remaining life expectancy. Almost every year, a smaller share of payments goes to those who are truly old and more likely to need assistance.
In sum, the recent widening gap in life expectancy, likely due to such factors as differential rates of cigarette smoking, deserves serious attention. But let’s not pretend that throwing money off the roof, or providing more years of retirement support to the non-disabled who make it to age 62, addresses the core issue. There are better ways to compensate than converting a system originally designed to protect the old into one offering middle-age retirement to everyone.
When the design of safety net programs is considered alongside that of our tax code, it is easy to see that our tax and transfer systems need to focus less on increasing consumption and more on promoting opportunity, work, saving, and education.
The government doesn’t affect work incentives just through direct taxes. Implicit taxes—that is, penalties for earning additional income—are everywhere, whether in TANF or SNAP, Medicaid or the new health exchange subsidy, PEP or Pease (reductions in tax allowances for personal exemptions and itemized deductions), Pell grants or student loans, child tax credits or earned income tax credits, unemployment compensation or workers compensation, or dozens of other programs. These implicit taxes combine with explicit taxes to create inefficient and often inequitable, certainly strange and anomalous, incentives for many households.
At some income levels, families face prohibitively high penalties for moving off assistance. Accepting a higher paying job could mean a steep cut in child care assistance for a single worker with children, for instance. For some, the rapid phaseout of benefits can offset or even more than offset additional take-home pay. Asset tests in means-tested programs create similar barriers to saving.
Not getting married is one way that people avoid some of these penalties or taxes and is the major tax shelter for low- and moderate-income households with children. Our tax and welfare system thus favors those who consider marriage an option—to be avoided when there are penalties and engaged when there are bonuses. The losers tend to be those who consider marriage a social or religious necessity.
The high rates and marriage penalties arising in these systems occur partly because of the piecemeal fashion in which they are considered. Efforts to design benefit packages more comprehensively could greatly improve both the incentives families face and the quality and choice of benefits they receive.
For more details, see my congressional testimony for today’s hearing on “Unintended Consequences: Is Government Effectively Addressing the Unemployment Crisis?” before the Committee on Oversight and Government Reform.