Adjusting measures of economic output for health

by on May 22, 2013 at 4:19 am in Data Source, Economics, Medicine | Permalink

This is the kind of argument which no one will successfully rebut, but no one really will take on and adopt either.  Does that mean we are defective?  Or is there simply ineffable wisdom in “how things have been done”?  Must we keep closed all Pandora’s boxes?

Here is the abstract from Mark L. Egan, Casey B. Mulligan, and Tomas J. Philipson:

Many national accounts of economic output and prosperity, such as gross domestic product (GDP) or net domestic product (NDP), offer an incomplete picture by ignoring, for example, the value of leisure, home production, and the value of health. Discussed shortcomings have focused on how unobserved dimensions affect GDP levels but not their cyclicality, which affects the measurement of the business cycle. This paper proposes new measures of the business cycle that incorporate monetized changes in health of the population. In particular, we incorporate in GDP the dollar value of mortality, treating it as depreciation in human capital analogous to how NDP measures treat depreciation of physical capital. We examine the macroeconomic fluctuations in the United States and globally during the past 50 years, taking into account how depreciation in health affects the cycle. Because mortality tends to be pro-cyclical, fluctuations in standard GDP measures are offset by monetized changes in health; booms are not as valuable as traditionally measured because of increased mortality, and recessions are not as bad because of reduced mortality. Consequently, we find that U.S. business cycle fluctuations appear milder than commonly measured and may even be reversed for the majority of “recessions” after accounting for the cyclicality of health. We find that adjusting for mortality reduces the measured U.S. business cycle volatility during the past 50 years by about 37% in the United States and 46% internationally. We discuss future research directions for more fully incorporating the cyclicality of unobserved health capital into standard output measurement.

The NBER link is here, does anyone know of an ungated copy?  Of course other forms of depreciation could be included as well (environmental?) and that too may smooth out business cycles, if we are willing to countenance such factors in the first place.

1 Ben Abbott May 22, 2013 at 4:44 am

Tyler, There’s a pdf link on the page you linked to (so its appears to be ungated)

2 bob May 22, 2013 at 6:01 am

Ben, you are probably reading this from a college campus network. From a general IP address, it is gated.

3 Jacob A. Geller May 23, 2013 at 8:00 pm

It is gated for me too (except when I log in via my institution).

4 dearieme May 22, 2013 at 5:37 am

“Because mortality tends to be pro-cyclical”: Mr Reagan’s joke was spot on, eh?

5 Andrew' May 22, 2013 at 5:45 am

All of his jokes were spot on. “I can’t recall” still kills me.

6 Andrew' May 22, 2013 at 5:44 am

1. Is the average ZMP worker doing anything harder/worse than what I’m doing? Hell no.
2. Everyone needs an “in case of Recession break glass for to-do list” box (e.g. Have kids)
3. Prepping is a good way to personally act counter-cyclically
4. Does nGDP targeting act upon expectations or on the actual supply of money?

7 Claudia May 22, 2013 at 6:42 am

“This is the kind of argument which no one will successfully rebut, but no one really will take on and adopt either.” Please. The national accounts will add investment in intellectual property (R&D and artistic originals) to GDP this year. Not saying this is easy. The physical capital stock including depreciation is imprecisely measured … surely human capital would be more difficult. But shifts in the productive emphasis of the economy have been reflected in our national accounting measures, so no reason to be so pessimistic. There is already a satellite account at BEA on health care to improve the quality of health measures.

On the paper, I was struck by mortality being pro-cyclical … suicide rates are counter-cyclical. Google to the rescue: here’s a paper by Ann Huff Stevens and coauthors that says the fluctuations in mortality are driven by elderly at nursing homes: … nursing home quality/staffing declines in economic booms and rises in busts. This is not to dismiss the findings in the post just to add to the interpretation.

8 Rahul May 22, 2013 at 6:48 am

What percent of mortality is suicides? It’s hard to imagine them having any significant effect on mortality’s cyclicity.

9 Claudia May 22, 2013 at 6:58 am

Of course it’s small, it was just that fact I had in my head. Plus I think of recessions as added stress / wear and tear on individuals, which would diminish health. But I was forgetting that overall mortality is driven by those out of the labor force. All very interesting.

10 Frederic Mari May 22, 2013 at 8:32 am

Thanks for doing the legwork, Claudia, that was my first reaction – “Mortality is pro-cyclical? Since when?”. But the nursing staffing level going down in good times makes sense.

More generally, I would be quite against including those measures of health in the national accounts. I think I’d be more amenable to introducing something for environmental degradation, (freely chosen) leisure and “stress”/quality of life.

Heightened mortality in nursing home is more a matter of health care choices and policy than GDP measurement…

11 Rahul May 22, 2013 at 8:48 am

Why are not environmental degradation etc. a matter of choice too then? I don’t see how you draw a distinction between some measures as choice and others not.

12 Frederic Mari May 22, 2013 at 9:35 am


First, I fully acknowledge that this is an a-priori. But I will try to explain my logic. GDP, esp. GDP per capita, is a way to try and measure how ‘wealthy’ a country/its inhabitants are, right? And usually, no matter what adjustments you make, ‘happy/development indexes’ end up strongly correlated to GDP. Being wealthy just make things so much easier kinda thing. Usually wealthier is also cleaner.

But the point of these indexes is to try to evaluate more completely the quality of life – wealth is a big chunk of your quality of life but other stuff matters. Environment (esp. from a sustainable pov) and health are two big obvious items.

Now, health can measured in a variety of ways. I think mortality at birth is a good measure but that’s going to be correlated to the GDP levels. ‘Stress’ would be most useful but evaluating that seems… difficult. Maybe suicide rates as a proxy? We should care about end-of-life quality of life so I am not against some kind of measurement such as maybe life expectancy for the 65+ but I am quite against a measure that turns out pro-cyclical for quirky reasons and thus mute the violence of the economic cycle and its real consequences on most people outside of end-of-life nursing homes…

13 Rahul May 22, 2013 at 10:34 am

Knowing that the fluctuations in mortality are driven by elderly at nursing homes, how realistic is it to use the $6.5Million dollar VSL (Value of a Statistical Life)?

More pertinently, if you look at the age-adjusted-mortality-value-correction (Fig. 2 in Egan’s paper) then there is hardly any difference between conventional GDP measures of recession and the mortality-adjusted GDP.

Much ado about nothing?

14 y81 May 22, 2013 at 6:45 am

Yes, and when you adjust for the value of leisure time, you find that we were actually better off during the Depression than before or after. Darn that Keynes for trying to end it!

Seriously, it’s a trivial exercise to find a defensible value for leisure time that does show a net welfare increase during the Depression. That’s why these attempts at “expanded GDP” are pointless and generally serve as attempts to cloak some particular political agenda in “scientific” disguise.

15 Frederic Mari May 22, 2013 at 8:37 am

You only need to use common sense (i.e. not buying into the neo-liberal marginalist ‘the markets are always clearing’ dogma) and define ‘leisure’ as normal people would i.e. leisure is freely chosen while unemployment/underemployment/disability are endured.

16 Alan May 22, 2013 at 8:08 am

Only hipster lefty greenies want to include factors like health, happiness and a sustainable environment. Listen, Tyler, if money can’t buy it, it ain’t worth nothing, no matter what fluffy-wuffy touch-feely metaphysical garbage they spout over their artisanal beers.

GMU needs an Inquisitor.

17 Ryan May 22, 2013 at 10:03 am

Only a hipster lefty greenie would try to tell me that money can’t buy happiness.

18 Alexei Sadeski May 22, 2013 at 11:06 am

Only a hipster lefty greenie would try to tell me that money can’t buy a good environment.

19 prior_approval May 22, 2013 at 12:05 pm

‘GMU needs an Inquisitor.’

Only when a donor steps up to pay for it – that is the GMU way.

But when they do, the sky’s the limit – as long as the checks don’t bounce, of course

20 Hazel Meade May 22, 2013 at 9:59 am

If we’re going to add health, why don’t we add happiness too?

Then, over time, we can gradually remove all that stupid production stuff and just focus on pure utility.
Can we have a measure of Gross Domestic Utility?

21 Rahul May 22, 2013 at 10:10 am

If Egan et al. want to adjust GDP for mortality, then why not include natality as well? Or is he already doing that somehow?

As an aside I wonder if natality exhibits any pro- or counter-cyclicity?

22 Brian Donohue May 22, 2013 at 11:04 am

If I recall correctly, births down, contraceptives up, during the Depression. I think people were having more sex too. Factor that in.

23 Jacob A. Geller May 23, 2013 at 8:03 pm
24 Ryan May 22, 2013 at 10:13 am

GDP is a measure of production (not utility, as some people seem to think, including possibly the authors given their liberal use of the term “value”). If it were a measure of utility we’d need some measure of consumer surplus, amongst other things.

NDP is a measure of production that takes into account the amount of capital consumed in production. From that standpoint, including certain forms of environmental depreciation make perfect sense (how much oil did we consume in the production of petroleum?). And from that same standpoint, it would make sense to include births in our measure of production, just as we currently include “investment” in physical capital.

25 ZHD May 22, 2013 at 11:42 am

Emailed. Cheers.

26 Frederick Harrison May 23, 2013 at 10:41 am

It looks like the economic crisis is hanging on tooth and nail. It’s my belief that our politicians need to take on the problems with more professionality. Maybe they need the services of the economic crisis specialists, the way certain counties have done in the USA. By hiring the Orlando Bisegna Index, they have resolved deficit problems and unemployment.

27 Jacob A. Geller May 23, 2013 at 8:05 pm

…GDP fetishism takes another blow to the head…

Comments on this entry are closed.

Previous post:

Next post: