Overselling faded dreams?

The April, 2017, issue of Science has a paper by Chetty, Grusky, Hell, Hendren, Manduca, and Narang on "The Fading American Dream" (ungated here). The paper documents falling income mobility. In particular, Chetty et al claim that, "the fraction of children earning more than their parents fell from 92% in the 1940 birth cohort to 50% in the 1984 birth cohort."

 There is something odd about Chetty et al's results. Women born in 1940 typically had children young, had fairly high fertility levels, and often did not participate in the labour force while they were raising children.  Many women in this generation had no earnings of their own. So how could they have earned more than their parents?  Given that women comprise half of the 1940 cohort, how could 90 percent of the cohort have earned more than their parents? 

As it turns out, despite the fact that Chetty et al use the term "earnings" throughout their paper, the focus of their analysis is, in fact, upon income:

Let y^k_ic denote the income of child i in birth cohort c, and let y^p_ic denote the income of his or her parents. In our baseline analysis, we measure income as pretax family income (summing income across spouses) at age 30. We measure incomes in 2014 dollars, adjusting for inflation using the CPI-U-RS. In sensitivity analyses (discussed below), we consider several variants of this income concept: using alternative price deflators, measuring income at age 40, measuring income after taxes and transfers, and adjusting for family size (emphasis added).

All of those 1940s women who were "earning" more than their parents? Many of them were "earning" their incomes through marriage.

One could argue - as indeed Chetty et al seem to do - that "earning" is synonymous with "income" or "standard of living":

When children are asked to assess their economic progress, they frequently compare their own standard of living to that of their parents... Such measures of “absolute income mobility”—the fraction of children earning or consuming more than their parents—are also often the focus of policy-makers when judging the degree of economic opportunity in the United States

Yet earnings are not the same as standard of living. To properly measure standard of living, income must be adjusted for family size, and also for taxes paid, and benefits received. A person born in 1940 might have 5 mouths to feed out of a $65,000 income, and pay face a 60 percent marginal tax rate. A person born in 1980 might have just one, and face significantly lower taxes.

In various sensitivity analyses, Chetty et al adjust for family size by dividing household income by the square root of the number of household members, and also consider inequality in after-tax incomes. It makes a difference to the results:

 Figure 3D shows that adjusting for family size considerably diminishes the estimated change in income mobility. Restricting the sample to fathers and sons increases the estimated mobility change.

Figure 3D shows something else: the dramatic decline in income mobility happened in the 1940s and 1950s. Children born in the 1940s and early 50s were able to out-earn their parents with relative ease because their parents were scarred, economically speaking, by the Great Depression and World War II. Focussing on the bottom line in Figure 3D (marked with triangles) shows that the earnings mobility of boys born in the early 1980s was not that different from the earnings mobility of boys born in the early 1960s. But I can see why Chetty et al pushed the 1940/1980 comparison.  "Earnings mobility declined dramatically in the 1940s and 50s" is not nearly as exciting as "income mobility is far lower in 1980 than it was in 1940". 

Likewise, I can understand why Chetty et al might prefer to talk about "earnings" rather than "income." Earnings are personal. To earn more than mom or dad is to be more professionally successful, at least by one metric, than mom or dad. Earning less than mom and dad means failing to live up to their standards. Hence people care deeply about their earnings relative to their parents. Earnings are clickbait-y in a way that "standard of living" or "adjusted pre-tax family income" is not. Given the deep and troubling social and economic changes that Chetty et al document, it would be understandable if they wanted to do everything possible to make people sit up and take notice of their results. Yet good intentions are no excuse for sloppy terminology.

The story of declining mobility is complex. Not everyone's prospects have been eroded to the same extent. Chetty et al conclude their article by saying: 

Our analysis yields two main results. First, children’s prospects of earning more than their parents have faded over the past half century in the United States. The fraction of children earning more than their parents fell from approximately 90% for children born in 1940 to around 50% for children entering the labor market today.

 Not children. Boys, perhaps, but not children. 

Comments

[Bob smith]

Great post.

[Ian Lee]

Superb analysis of the, er "fading methodological rigour". And by a stroke of serendipity, Frances' excellent post follows Doug Saunders' flawed Globe piece last week, "Can we ever knock down the walls of the wealthy ghetto?" where he clearly suggests that people in the top 1% (over $220K/year) do not earn their successes on their merits - due to the "closed doors" of the "ghettoized wealthy".

The idea of a lack of intergenerational mobility due to "closed doors" - using an income cutoff of $220K/year - is curious as this income standard captures most or all university Presidents, a significant number of university Provosts and some Deans, a goodly number of full professors at least in Ontario (per university sunshine lists), most Canadian hospital presidents, most medical doctors in Canada, many federal and provincial deputy ministers such as the Chief Statistician of Canada, some NGO Presidents (per the recent Globe & Mail list), most legal and accounting partners, the Governor General of Canada, many superior court justices including the Chief Justice of Canada, not to ignore Prime Ministers and Premiers, some cabinet ministers and of course FP 500 executives.

I hope someone unpacks these issues. Is the issue we should focus on a lack of intergenerational mobility? or children under performing their parents? or an inequality issue where it seems that wealth and income are sometimes conflated or earnings with standard of living as Frances notes?

Yet, while admittedly anecdotal, in numerous Globe & Mail or CBC profiles over the years of some distinguished Canadian thought leaders, a very common theme is the very modest - if not downright poor or working class - family circumstances of the future 1 percenter. See the recent profile of Chief Justice McLachlin or the profile of Governor Mark Carney growing up north of 60 in the home of a high school teacher.

Restated, as the number of wage earners in the top 1% increases, does this not suggest a more meritorious and more mobile society, for the occupations and professions cited in the top 1% require advanced degrees in difficult disciplines? Surely a PhD in mathematics and eventual Chief Statistician of Canada - after arriving from communist Hungary (Felligi) - cannot be attributed to private schools and a lack of inheritance tax in Canada?

[Frances Woolley]

Ian, Bob, thanks for your comments and kind words about the post.

[cm]

Perhaps tangential but I think on point: (material) "standard of living" and "quality of life" are frequently conflated if not equated, but they are different things. "Professional"/"white collar" work has become increasingly tougher, less secure, and more taxing in terms of hours, 24/7 availability, etc. (we now have laptops, handheld devices, mobile internet, and the thereby enabled expectations of being online). In general, job security and security of livelihood has declined in aggregate. Increased income doesn't really compensate for it - what use is the money when there is less (quality) time to enjoy it, and higher incomes just get priced into "required" spending?

My usual thought experiment is to disregard money and look at the flow of goods and services people receive, and what they are asked to give in return (e.g. aside from money, "wasted" time and attention).