Mark Perry and I have received (and continue to receive) a great deal of push-back on our claim, published recently in the Wall Street Journal, that America’s middle-class has not stagnated economically over the past three or four decades. In the blogsophere, Paul Krugman, among others, challenges our claim. In private e-mails to Mark and me, the eminent Washington Post reporter Thomas Byrne Edsall adds his voice to the challengers.
It’s trite – but true enough to make it worthwhile to say – that for any coverage of this topic to have any claim of completeness it would require far more space than is allotted in an op-ed. And far more space also than whatever additions Mark and I will offer in follow-on blog posts to that op-ed. Here’s just a partial list of the issues that must be addressed by any comprehensive attempt to measure the course of middle-class living standards since the 1970s:
- How well do adjustments for inflation (such as the CPI) work – and, especially, how well do they account for changes in product and service quality?
- How well do changes over time in the measures of various features of aggregate statistical groups (e.g., “amount spent on food by households in the middle income quintile, 1980-2012″) measure changes in the corresponding feature of individual, flesh-and-blood people who were – for part or all of the relevant time period – included in that statistical group (e.g., “amount spent on food by the Mark Perry household of Flint, MI, 1980-2012″)?;
- Care must be taken (by all ‘sides’ in this debate) when measuring consumption quantities according to the amounts or the proportion of real income spent on various consumption goods. Suppose a statistician who died in, say, 1973 is resurrected today and immediately hurried to the task of measuring Americans’ ability to communicate today as compared to the mid-1970s. Being unaware of the huge change in telecommunications products over the past 40 years, that statistician might conclude that Americans today are far worse off on at least one dimension of communications: talking in real time to people who are more than 20 or 30 miles away. The reason for his dreary conclusion is that the amounts that Americans spend today buying long-distance telephone calls within the United States has plummeted, both absolutely (in dollar terms) and, of course, also as a portion of household expenditures. But obviously such a conclusion by this statistician would be a great mistake.
Consumption expenditures change with changes in the prices of the goods and services consumed or with changes in the quantities consumed (or both). A once-scarce and high-priced service (long-distance telephone communications) that is now super-abundant at the margin (and, hence, priced at $0) shows up as being unconsumed if the statistician looks only at the amounts of money spent on the service.
- The complexity noted just above gets even deeper. Suppose that the price of good X falls significantly (but not to $0); say, the real price of X falls by 80% (from $100 to $20). Further suppose that people choose, as a result of this price decline, to buy, each period, 500% more more of X. Previously people bought 10 units of X (spending a total of $1000 on X each period). Now, because of the fall in the price of X, people voluntarily buy 60 units of X each period (spending now a total of $1,200 on X each period). A naive glance at the statistics on consumption expenditures might lead a careless observer to conclude that consumers are now worse off as regards their access to X. As compared to the past, consumers are spending more absolutely (and perhaps also as a proportion of their incomes) on X. Looking only at (inflation-adjusted) monetary amounts, or income proportions, spent on some good or service (or some bundle of goods and services) has its value, but – as this simple example shows – also its dangers.
- Relatedly, suppose that the quality of X improves dramatically over time, even while its real price rises. (Or suppose that what is classified as X expands over time.) Greater real expenditures (either absolutely, as a proportion of income, or both) on X tell us very little about any resulting changes in consumer well-being.
Suppose that in 1980 the only form of health-care available was band-aids. No MDs, no dentists, no hospitals, no antibiotics, no anything except band-aids. Fast-forward to 2013, a time during which health-care includes all that we denizens of 2013 actually have access to – band-aids; primary-care MDs; pediatric gastroenterologists; MRI machines; statins; antidepressants; lasik eye surgery; erectile-dysfunction medications; on and on and on. A statistician finds that people in 2013 spend much more on health-care than they did in 1980. Is it proper to conclude from this correct statistical finding that the rise in health-care expenditures – this rise in health-care costs – is a problem, a curse, something that we should lament? Suppose that between 1980 and 2013 the portion of household income spent on non-health-care basics (e.g., food, clothing, transportation) fell – a fall that is offset by the rise in the proportion of household income spent on health care. Is it correct to conclude that, because the portion of middle-class household income spent in 2013 on the sum of non-health-care basics plus on health care is unchanged from the portion spent on these items in 1980 that middle-class households are no better off economically today than they were in 1980?
- While Mark and I do touch in our essay upon the question of inequality, the bulk of it – and the most important thesis that we wish to defend – has much less to do with inequality than with the absolute economic well-being of middle-class Americans. That is, even if we were to grant that both income and consumption inequality has risen over the past few decades, that fact alone says nothing about the absolute economic well-being of middle-income and poor Americans. While (I believe I speak here for Mark) we believe that consumption inequality has in fact declined, our larger, more central, and most important point is that middle-class Americans are today far better off economically than they were 30 or 40 years ago, regardless of how their well-being today compares to that of rich Americans.
- Finally, I repeat that how one comes down on this question should reveal little about one’s ideological priors. I’ll say no more about this matter here, except to ask “Progressives” who insist that America’s middle-classes have stagnated economically since the 1970s: Do Social Security, Medicare, the Americans with Disabilities Act, Food Stamps, the Departments of Energy and of Transportation and of Education and of on-and-on, NSF and NEA funding, Fannie Mae and Freddie Mac and the FHA, the SEC, the EPA, and all of the many other jewels of FDR’s New Deal and of LBJ’s Great Society (and of other eras as well, including not-insignificantly the presidency of Richard Nixon) really work so poorly and so delicately that the many blessings that they promise have been completely nullified for most Americans by the marginal tax-rate reductions and relatively modest deregulations of the Carter-Reagan years?