« Links for 02-21-16 | Main | Links for 02-22-16 »

Sunday, February 21, 2016

'Absolute Changes in Living Standards'

I thought this was interesting:

Absolute Changes in Living Standards, by Dietrich Vollrath: This may or may not be interesting, I'm not entirely sure myself. But I got a question from an undergrad in my economic growth course, and it was one of those questions that kind of catches you off guard because you hadn't really thought about it. The question was - paraphrasing - "Why do we look at growth rates rather than how much GDP per capita actually changed?". ...
Why do we focus on growth rates - percentage changes - as opposed to absolute changes in GDP per capita? ... For example, a growth rate of 2% implies an absolute change of 1000 in an economy with a GDP per capita of 50,000, and only a change of 200 in an economy with GDP per capita of 10,000.
First, what does the distribution of absolute changes in GDP per capita even look like? I had no idea. So I calculated the absolute growth in real GDP per capita per year for each country in the PWT. ...
As you can see, for India ... absolute growth tended to be very small. On average, from 1950 to 2011, GDP per capita rose by 51 dollars per year in India. The median was 42 dollars. Despite these small absolute changes, the percentage growth rate in could actually be quite high because GDP per capita is very low to begin with.
Compare that to the United States... For the US, average absolute growth in GDP per capita was 482 dollars, with a median of 617 dollars. Of course, those large changes in GDP per capita reflect very small percentage growth rates in the US, because GDP per capita is already so large.
From the perspective of growth rates, the US isn't growing very quickly compared to India, at least over the last 25 years. But in absolute terms, the increase in living standards in the US is about 10 times higher than in India in a given year.
Should we care about these absolute changes? ... It is really common to assume "log utility"... Log utility has the very particular property that the same growth rate of GDP per capita yields the same absolute change in welfare. If India and the US both have growth rates of 3%, then welfare in both countries rises by 3%, even though the absolute change in GDP per capita in the US is much larger. ...
Now what if utility is "more curved" than log? That is, what if marginal utility falls very quickly as GDP per capita goes up? Then ... the richer country would have a smaller change in welfare than the poorer country. ... But we could go the other direction, and look at utility that is "less curved" than log, or where marginal utility diminishes very slowly. Now,... the richer country has a higher gain in welfare. ...
An argument for thinking of utility as "more curved" than log is subsistence constraints. ... On the other hand, habit formation would be a force pushing towards "less curved" utility. ...
Ultimately, then, I think the answer to the students question is that we think there is diminishing marginal utility, and so growth rate rates make sense because they tell us something about absolute changes in welfare. But...who said utility of GDP per capita has to be as curved as log? ...

    Posted by on Sunday, February 21, 2016 at 04:03 PM in Economics | Permalink  Comments (68)


    Comments

    Feed You can follow this conversation by subscribing to the comment feed for this post.