Born Lucky, Live Long?

The Washington Post reports today that a Dutch study has found:

“People who happen to be born in good economic times live longer than those who aren’t and the apparent payoff from an economically fortuitous birthday lasts a lifetime, according to a team of European researchers.”

Um, no.

I haven’t seen the raw data from the study, but I did read through much of the actual IZA report. There appears to be some glaring potential flaws in this study, the most notable of which is this: Is there a causal relationship between economic expansions and mortality? Or are these merely two unrelated issues occurring contemporaneously?

I can pair all kinds of unrelated phenomena with a positive correlation, but that doesn’t mean that one causes the other. Imagine a group of school superintendents in California noticing that when Las Vegas gambling receipts rise, school budgets go also go. Should they try to increase Las Vegas tourism as a method of improving their own budgetary situations — or, is the same underlying factor responsible for both? It’s more likely that an expanding economy both improves state budgets for schools and enhances people’s abilities to take vacations.

Applying the same logic to the 19th century Dutch study: Is it possible that epidemics had a negative impact on both mortality and economic expansions? One would imagine that productivity slows and the sale of goods and services decreases during widespread health crises. (You don’t consume or produce much while dead or incapacitated). How much did people’s fear of contracting a serious or mortal disease during infectious outbreaks curtail their economic activity?

One can even imagine conversations like this:

“What? Go to the market today? Are you daft, man? Don’t you know there is an out break of cholera/smallpox/influenza/fill in the blank?”

I suspect the better question is this: “Did widespread epidemics in the 19th century cause both an increase in mortality and decrease in economic activity?”

Some studies (See Prof. Chris Ruhm of UNC) conclude the exact opposite relationships — i.e., good times make you sick. I haven’t correlated these studies either, but I am always wary when two extremely complex phenomena are “demonstrated” to have a causal relationships. Life tends to be more complex than that.

The GOP Problem With Women (Born Lucky, Live Long)
Richard Morin
Washington Post, Sunday, January 11, 2004; Page B05

See also Department of Economics at the University of North Carolina at Greensboro

“Good Times Make You Sick”, Journal of Health Economics, Vol. 24, No. 4, July 2003, pp. 637-658.
“Does Drinking Really Decrease in Bad Times?,” Journal of Health Economics, Vol. 21, No. 4, pp. 659-678.
“Are Recessions Good For Your Health?,” Quarterly Journal of Economics, Vol. 115, No. 2, May 2000, pp. 617-650.
“Deaths Rise in Good Economic Times: Evidence From the OECD”

Hat tip to Marginal Revolution

What's been said:

Discussions found on the web:
  1. Terry commented on Jan 13

    The quote from WaPo at the beginning suggests the study relates mortality to economic conditions only at the time of birth (not to average economic conditions over the individual’s entire life. If this is so the entire thrust of your argument collapses. Think about it.

  2. Barry Ritholtz commented on Jan 13

    I do not disagree with the article’s statement, that it is “particularly useful [from a policy standpoint] to focus on children aged between 1 and 7 in bad economic conditions.”

    However, I think the study used a faulty analytical methodology to reach that conclusion, one that is not statistically sound,

    Being “born during an economic expansion,” according to the study, added “nearly three years to the life of the average respondent.” Is that the benefits of a healthy childhood (age 1-7) — just 3 years? Exercise, eat your vegetables, don’t smoke, go to the Doctor — and you’ll live 3 years longer? ; )

    More significantly, the study compares the lifespans of 3000 various people born between 1812 and 1912. It then determines whether they were born during a recessionary or expansionary period. Then it compares their lifespans. It is obviously foolish to compare the lifespan of someone born in the early 19th century with someone else born in the 20th Century. Given the advances in medical technology 50 and 100 years later (i.e., we cannot contast 1812, 1861 AND 1912 births), they are incomparable.

    Who is going to have a longer average lifespan, 1812 expansion babies, or a 1912 recession babies?

    So the study compared longevity of the 3000 “random” individuals against the rest of the population, before and after their births.

    That brings me back to the original issue: isn’t possible — even likely — that the same forces which caused the increased mortality rates (shorter average lifespan) impacted economic conditions? Remember, we’re dealing with average population lifespans, based on their period of birth. Any kind of a broad epidemic (with significant mortality increases) would lower the average lifespan a few years of that age bracket. I.e., the 1850-55 births, if there was a major epidemic going on, would have a shorter average lifespan than the 1856-60 group, if there wasn’t a health crisis going on then.

    That same epidemiology would also slow down consumption and production. Which causes which? I am suggesting that regular minor epidemics (influenza) as well as more major health problems (small pox, TB) are a major underlying cause of both lifespan and economic activity.

    What the study found, in my opinion, was a coincidental relationship between the two, and not a causal one.

  3. thebizofknowledge commented on Aug 12

    I stumbled across your blog while I was doing some online research. You make an excellent point; correlational relationships are often reported in the media as causal relationships, and, thus, the public is, once again, misinformed.

Posted Under