Bryan Caplan  

Against the Human Development Index

Fuel Economy and Relative Pric... Two Disagreeable Questions for...
The authors of the "U.S. Unemployment Rate Now as High as Europe" report have turned down my bet.  One of the authors did however counter-offer a bet that Scandinavia's scores on the Human Development Index would beat America's at the end of the Obama administration.  That's funny, because I was already planning to blog against the usefulness of the HDI today, provoked by the CEPR report's remark that:
The case for the superiority of the U.S. model was always exaggerated.  For one thing, it tended to ignore the relatively lower performance of the U.S. on broader quality-of-life measures like the Human Development Index.
Now what exactly is the HDI?  The one-line explanation is that it gives "equal weights" to GDP per capita, life expectancy, and education.  But it's more complicated than that, because scores on each of the three measures are bounded between 0 and 1.  This effectively means that a country of immortals with infinite per-capita GDP would get a score of .666 (lower than South Africa and Tajikistan) if its population were illiterate and never went to school.

So what are the main problems with the HDI?

1. I can see giving equal weights to GDP per capita and life expectancy.  But education?  As a professor and a snob, I understand the appeal (though a measure of opera consumption would be even better).  But in terms of the actual if not professed values of normal human beings, televisions and cars are a lot more important than books.

2. When you take a closer look at the HDI's education measure, it's especially bogus.   2/3rds of the weight comes from the literacy rate.  At least that's not ridiculous.  But the other 1/3 comes from the Gross Enrollment Index - the fraction of the population enrolled in primary, secondary, or tertiary education.  OK, I feel a reductio ad absurdum coming on.  To max out your education score, you have to turn 100% of your population into students!

3. The HDI purportedly gives equal weights to three different outcomes, but bounding the results between 0 and 1 builds in a massive bias against GDP.  GDP per capita has grown fantastically during the last two centuries, and will continue to do so.  In reality, there's plenty of room left for further improvement even in rich countries.  But the HDI doesn't allow this.  Since rich countries are already close to the upper bound, the HDI effectively defines their future progress on this dimension out of existence. 

To a lesser extent, the same goes for life expectancy: While it's roughly doubled over the last two centuries, dying at 85 is not, contrary to the HDI, approximately equal in value to immortality.

The clear winners from this weighting scheme, of course, are the literacy and enrollment measures, both of which have upper bounds that are imposed by logic rather than fiat.

4.  The ultimate problem with the HDI, though, is lack of ambition.  It effectively proclaims an "end of history" where Scandinavia is the pinnacle of human achievement.  Admittedly, I've never visited Scandinavia.  But when I see it for the first time this August, I'm pretty sure I won't say to myself, "Wow, it can't get any better than this!"

At this point, you might ask, "Yes, but will you take the HDI bet, Bryan?"  Nope.  Scandinavia comes out on top according to the HDI because the HDI is basically a measure of how Scandinavian your country is.  While Obama is moving us in that direction, I don't think he's going to be able to take us all the way there.

Comments and Sharing

COMMENTS (39 to date)
kingstu writes:

"Scandinavia comes out on top according to the HDI because the HDI is basically a measure of how Scandinavian your country is."

This might possibly be the funniest thing ever written.

Jonathan Bydlak writes:

I second kingstu.

RL writes:

Actually, I liked the line "This effectively means that a country of immortals with infinite per-capita GDP would get a score of .666 (lower than South Africa and Tajikistan) if its population were illiterate and never went to school." even better.

You're a great teacher, Bryan.

MCR writes:

You might have to qualify your comment about the gross enrollment rate.

As I recall, the gross enrollment rate has all people enrolled in school in the numeratior and the number of school age individuals in the denominator.

As a result, gross enrollment rate can exceed 100% if lots of adults go back to school.

Peter Twieg writes:

That is a great take-home quote.. however, it seems like this argument could be taken a little further to say that even if the weightings did reflect an acceptable metric of average (material) quality of life in a country, that it's arguable that this does not imply that we should simply adopt policies that would reliably maximize our HDI rank. Most notably, what factors determine the growth in HDI over time, and do the Scandanavian countries actually promote these things as well... or perhaps, as seems plausible to me, they benefit from a lot of innovation and knowledge spillovers that wouldn't exist if governments ham-fistedly were trying to maximize HDI rank?

Of course, this runs into the issue of upper bounds again - if a country at the top of the HDI listing improved its life expectancy by five years this wouldn't actually improve its score, it would just reduce everyone else's.

Franklin Harris writes:

I third kingstu.

J Cortez writes:

I second the country of immortals with infinite per-capita GDP as a classic line.

I have trouble accepting stuff like HDI, the Gini coefficient and even crazier things like a "happiness" index. The people doing this appear to be completely unaware of individual human nature, cultural biases and sometimes even reality. Not everything in this world can be equated with statistics, regressions and mathematical models.

The Snob writes:


But it gets worse. AFAICT, the education component boils down to adult literacy plus school attendance.

Adult literacy correlates highly with income, so this is basically just an excuse to count education twice. Meanwhile, the "gross enrolment ratio" does nothing to control for quality. Advanced study in a madrassa counts for the same as medical school.

Felix writes:

Why, oh why, do these various metrics even include 19th century measures?

Life span? It's the same for all the "developed" areas of the world.

Education level? It's the same for all the "developed" areas of the world.

Infant mortality? It's the same for all the "developed" areas of the world.


The differences between the "developed" areas of the world in all of these measures is noise and is probably dominated by differences in measurement techniques, etc.

Other examples of like-Scandinavia thinking: Opinion pieces dealing with health care. They are often, "This health care system is best on a scale of how much like this health care system a health care system is." Or, "This health care system is best because it's tops on the measures it uses to measure how well it does."

That sort of thinking can get fun if you note that the US health care system is clearly the best in the world. How do we know this? Because the US spends the most on health care. If it were such a lousy system, people wouldn't be willing to spend so much money on it!

Michael Makowsky writes:

This was funny. It also represents one of the biggest difficulties when trying to run regressions using indices as independent variables, something which I am guilty of. It can be awfully difficult to figure what is driving the coefficients, and what the magnitudes really mean.

Don't get me wrong, though. I hear the Star Trek utopia of the next millennium is loosely based on the Scandinavia of next week.

Kurbla writes:

Indexes as HDI are not meant 'to work' in all possible worlds, but in existing world, comparing relatively similar countries. One can ridicule GDP alone on the same way, imagining variety of dystopian societies with very high GDP but very low quality of life. However, even if you prefer GDP only, Scandinavian countries appear to have higher GDP/capita than USA.

jsalvati writes:

Could you please comment on this:

John Quiggin claims that more recent evidence does not show an effect on average levels of unemployment from employment protections.

Tom Ault writes:

GDP Per Capita, PPP, 2008, according to the IMF

Country GDP (PPP, USD)
Denmark $37,265.79
Norway $53,450.67
Sweeden $37,245.09
United States $46,859.06

Of the three Scandinavian countries, only Norway has a higher per-capita GDP, and that's only because of its oil exports. Sweeden and Dennmark have significantly lower GDP per capita. The Scandinavian countries only have higher per-capita GDP if you use nominal figures as opposed to ones that correct (however imperfectly) for differences in price levels between countries. Nominal GDP figures are subject to fluctuations in the exchange rate that have nothing to do with standard of living within the respective countries and contributed signifcantly to the rise in nominal (but not real) GDP per capita of Denmark and Sweeden.

BlackSheep writes:

jsvalti: Bryan just blogged on that -- it covers everything mentioned by Quiggin:

My 2 cents if you allow me. I'm an autodidact in economics, but from what I understand flexible labor seems clearly superior because it doesn't rule out rigid contracting. Employers have the correct incentive to evaluate the trade-off employees are willing to take between rigidity and compensation when they open positions. They afterall are able to get the producer surplus between the cost of producing job security and the willingness to forgo pay relatively to the currently practiced baseline.

Matt writes:

Yeah, Bryan completely misinterpreted the gross enrollment index. It's the number of persons in primary, secondary, and tertiary education divided by the total number of school-aged persons (each component separately, I believe). So, to max out your score, you just need to make sure all school-aged children are in school. Tertiary education is probably not as essentially, but who wants to argue with keeping kids in school until they graduate high school?

Snark writes:

A genuine Caplan moment (applaudere)!

Scandinavia comes out on top according to the HDI because the HDI is basically a measure of how Scandinavian your country is.

Yes, the HDI is fundmentally flawed. What's needed is a more accurate and objective measure of human well-being to satisfy all parties to the bet. Perhaps we should consider the Happy Planet Index (HPI), where the U.S. currently resides in last place.

Snark writes:

...where the U.S. currently resides in last place.

Well...last place relative to EU members. We are only marginally better off than some of the poorest African countries.

eccdogg writes:

Perhaps the best metric for the bet would be out migration as a percent of total population.

The question then is what percent of people who are in country x choose to leave it. Probably not a great comparison for the third world where leaving may be difficult, but a great comparison for rich countries.

If a country has such a great system you would suspect most people to want to stay.

The Snob writes:
Tertiary education is probably not as essentially, but who wants to argue with keeping kids in school until they graduate high school?

I'll bite. There is such a thing as over-investment in education. The Amish manage to do pretty well with their kids only getting K-8 and they have a higher standard of living and more technological sophistication than most of the developing world.

My take in these places would be to focus money on basic literacy, math, and English, and then work on getting cheap high-speed internet service to the people. A big part of "education" is simply introducing people to the information that is available and the cost of doing that is dropping rapidly.

Here in the US I wonder how many people graduate high school and then go on to work dead-end jobs they were perfectly qualified for at the age of 14 or 16. Some kids (disproportionately female) enjoy the classroom environment and get a lot out of it.

Some kids (disproportionately male) find it stifling and withdraw or rebel. Some of this is correlated with cognitive ability, a lot of it is just personality. In the best case they grin and bear it and then find their path in adult life. In the worst case, it actively damages them. Those who fail to conform to the system's expectations are drugged with anti-depressants, fed into the criminal justice system with zero-tolerance policies, and exiled as dropouts.

On top of all this, the money that goes into our secondary school system, especially in the cities where the problems are most acute, is eye-popping.

Ben writes:

I'm afraid I don't have anything useful to add to Caplan's excellent article, but I have added bits of it to the Wikipedia article.

Conor writes:

The problem with non-GDP measures of economic well-being is that they are designed by people with a strong anti-market bias. The unintentionally hilarious Happy Planet Index mentioned above is an extreme example of this phenomenon, but the difference between that and the HDI or Human Poverty Index is one of degree, not kind.

Bryan makes excellent criticisms of the HDI, but what's missing is an explanation of WHY the HDI is flawed the way it is. I assure you, to the statist designers and promoters of the HDI, it's functioning perfectly.

Snark writes:


The unintentionally hilarious Happy Planet Index mentioned above is an extreme example of this phenomenon

The hilarity was most intentional.

Mike Rulle writes:

I understand the desire to respond rationally to ridiculous measures like the "HDI". However, it is a waste of time to do so. When the answer is already known (e.g., "the US way of life is bad"), there will always be a measure to "prove it". I would prefer to counter these with equally absurd "indices", demonstrating how the US superiority has been under stated.

Remember the parody paper written 10-15 years ago by a physics professor that "deconstructed" various scientific principles? It was published as serious by some transgressive publication. "HDI" reminds me of that. Indices like the "HDI" are so absurd, it is as if nihilists or "parodists" created them. However, it is more likely very dumb and earnest people did so.

LemmusLemmus writes:


if the HDI "reminds" you of "Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity", it just goes to show you haven't read that paper.

Felix writes:

eccdogg writes:

Perhaps the best metric for the bet would be out migration as a percent of total population.

Just so eccdogg doesn't get the impression that no one read his comment, I vote it to the top.

A related measure, oddly not mentioned here yet, is, "What do people pay to live in the place?" Since prices are oft times locally relative, this can be a muddy measure, but a population in/out movement measurement has the same problem.

Relative measures: Are there cases of people in A wanting to move to B. People in B wanting to move to C. And people in C wanting to move to A? In total, that is.

All of these measures of goodness created by individuals and interest-groups are most useful to people who rely on their survey to accurately tell them whether their innovative product will be a success.

Barkley Rosser writes:

Well, the US does pretty well on the education index, and as the numbers have been shown, also does well on PPP GDP per capita, which is what goes in the index, not nominal GDP per capita. A good third of the comments here were made by severely ignorant people, but I shall not go through the list of inanities, most of them decrying the index on ideological grounds for being ideological. Gag.

So, what pushes the US down? Oh, it's that snarky life expectancy number, where we are about 32nd in the world. Yeah, sure, we are only a few years shorter-lived than those Japanese and long-lived Nordics, and, heck, we know that whining about health care when we pay a higher percent of our GDP for it than any other country on the planet is a socialist plot because of course as a free market country with a more privatized health care system than other countries we are more efficient, even though Hayek was for national health insurance, but, well...

H. J writes:


Do you honestly believe life expectancy is more an indicator of successful health-care systems than it is an indication of social behaviors? The last time I checked, our partially-socialized health-care system did not cause our higher rates of obesity and deaths unrelated to health (e.g. homicide, car accidents, ect).


Joshua Holmes writes:

US life expectancy is lower but not significantly. The US has a murder rate 9 times that of Japan, 7.5 times that of Norway, 6 times that of Denmark, 3 and a third times that of France, etc. That is a serious problem that needs addressing, but it's not a serious health care problem per se. Stats are at Wikipedia's list of countries by homicide rate.

Also consider that the US has a much higher traffic fatality rate per capita than these other countries: 2.5 times that of Japan, 2.6 times that of Norway, twice that of Denmark, 1 and two-thirds times that of France, etc. Again, this is a problem but not a health care problem. Stats are here.

Regress those two to the OECD mean, and we look a lot better for life expectancy.

Jason Malloy writes:

The relationships between the measures that make up the HDI are so tight that it doesn't matter much either way: correlation between HDI rank and GDP rank = .95.

Barkley Rosser writes:

Hmmm. Tried to post earlier, but did not take. Here goes again.


Sure, high correlation, but if it does not matter, then why all the fuss? Because it does matter, and the fuss is that the US is really way up there on real GDP per capita but then dragged down lower, below all those naughty Nordic economies (eeeeek!). Hence the fuss.

JH and HJ,

So, maybe part of the problem is not just that we pay more for a health system that does not deliver so well, maybe its delivery is about the same as some others, but all this obesity and homicides, and so on we have. Well, but is not the issue whether or not life expectancy should enter into the HDI strongly? Bryan Caplan actually agrees that it should, getting more worked up about the education part. Does not the fact that the US has a lot more fat murderers than many other countries indicate a lower level of human development?

BTW, in 2008 life expectancy in Sweden was 80.68, tenth among the 221 world "entities," while the US's was 78.06, 45th among the entities, a more than two and a half year difference. Is that "insignificant"?

Martin Rytter writes:

Coming from Denmark I can't help myself throwing in a few problems we have:

Most people spend years on "free" education that they never use.

People choose the education they think would be most fun and not the one that would help them pay for their own lifes. For instance we have a shortage of engineers and teachers while we have too many musicians, masters of history and litterature.

39% of the working age individuals (16-64) recieved some form of social payment (permament of temporary) in 2007. Transfers to students not included (another 7% points as far as I remember)!

Even though our official pension age is 64 many people stop working when they turn 60. Why? We have a special "pre pension" for the "people who have had a hard and long working life". This transfer is designed so that anyone can apply for it a get it. You just have to submit a form.

Smart people don't want to move to Denmark because of high taxes. Smart Danes immigrate. Goverments solution: Have a special low tax rate for high wage foreigners moving to Denmark the first years they live here. Normal tax-paying danes don't even complaint about that! Btw. this special tax is very popular among sport stars moving here!

Incentives works beautifully - even in Denmark.

I'll propably immigrate when the state (we don't speak of tax-payers here) has paid my PhD in software engineering... I just have to convince my girl friend. The state currently pays for her medicine studies.


Barkley Rosser writes:


Ummm, first of all, apparently you want to "emigrate" not "immigrate," unless you are talking in regard to the US.

Clearly for someone with the potential to earn a high income, which would be he case for a Ph.D software engineer, you will probably be much better off emigrating out of Denmark and to some location in the US. You will shortly be in the top couple of percent of US income earners and will certainly have access to the best of our medical system through a really good insurance plan, or be able to pay for it yourself. The problems with the US health care system show up for the 40 million or so who are not poor enough to qualify for state-supported medicaid, but not in a good enough job or well enough off to get any medical insurance (or have a lousy and expensive plan they cannot afford).

So, fine. You are better off leaving Denmark, and maybe more of your compatriots would be better off studying engineering or computer software than music or history. However, studies continue to put Denmark on top of the list in the world for overall reported happiness, which apparently does not include you (maybe this is why your girl friend still needs convincing?), and you will have a sympathetic ear on this blog with Bryan Caplan, who does not approve of or believe in all those darned happiness studies.

scipio writes:

@Barkley Rosser

Being from Denmark I will offer some further information. This might be interesting since this seems to be the OPTIMAL developement direction ;-)

Not all welleducated people wants to leave Denmark. I have a master in Computer Science and a Ph.D. in Math., and I am planning to stay. I could certaintly earn more (post tax) in US than in Denmark, but the picture is much more complicated.

In Denmark the "deal" is that I pay a lot of tax but also get a number of "services" from the state. You will probably argue that I could get the same services by paying privately for it, but I get something more out of my tax: The state helps people in need. That helps my concience AND reduces the crime level significantly.

It is correct that Denmark "routinely" end at the top of the happiness indexes. There are probably a lot of historical reasons for this. The state does however provide free medical service (also if you do not work) and that probably makes people less worried about the future.

Barkley Rosser writes:


Fair enough, and I am not unsympathetic at all here, in comparison with probably most other readers. After all, what set this off is that Bryan Caplan is finally going to Scandinavia, Sweden in particular I think, and is just itching to find something critical to say so that he can diss those countries with their awful high taxes (eeeeek!), starting with trying to knock down all these supposedly ideologically biased international measures such as the awful Human Development Index (which does not give the US enough credit for its people being free to eat too much and have guns easily so that they can kill each other left and right, including themselves and puts too much emphasis!).

Which brings me to a serious poke back, actually one that is one of the most standard of cliche criticims of the Nordic nations: their relatively high suicide rates, although I think Denmark does better than some others, such as Finland. I realize that this may be due to seasonal affect disorder, but still, it does seem to undercut all those happiness studies results somewhat.

I have commented here before on happiness studies, which I see gobs of because of the journal I edit (JEBO). The least meaningful ones are exactly the sort we are talking about here: cross-section studies across countries. Not only are we comparing across people, but across cultures, and I have observed very different attitudes in different countries about what people say about themselves. I can list more than one country where people love to whine and complain, but really do not seem all that unhappy (and are not killing themselves in large numbers), indeed would look upon a compatriot who went around babbling about how happy they are as some fool beneath contempt, an idiot not worth talking to. My on-site observation in Nordic countries is of a certain conformism in the form of not wishing to stick out or complain, which would fit in with people reporting that they are happy to a questioner, even if they are planning to go home and do themselves in with an overdose of aquavit.

So, I am viewing all parties here a bit skeptically, prepared to question the claim that the Danes are the happiest in the world (indeed finding the question somewhat meaningless), while at the same time finding a certain amount of comedy in Bryan's exaggerated efforts to find something, anything, wrong with the HDI or other indices that show the Nordic countries doing pretty well. Indeed, on most of those freedom indices he and his pals love, those countries end up doing pretty well, with only those high taxes being the negative (but in the US we have a political party that treats opposition to taxes as a sacred duty and party line that must be kowtowed to by one and all, hack, cough).

scipio writes:

@ Barkley Rosser

Hmm, these happiness studies are far from my area and I do not know how reliable they are ...

There is another series of studies regarding how much people trust in strangers, and in Scandinavia two out of tree trust in strangers. Furthermore all of the Scandinavian countries are fairly "safe". I do not say that that is a guarantee for a happy life, but it does remove some of the uncertainties.

Scandinavia is very different from the US, high taxes etc. and I am certain that Bryan Caplan will find a lot to dislike. On the other hand Scandinavia could be an inspiration of how society could be organized in a very different way.

Joel writes:


I've been to Scandinavia: it’s pleasant but not essential. It's a (mostly) ethnically heterogeneous society that values stability and comfort over individual freedom (unless that freedom relates to sex).

Like Germany, both have because the locals won't do the jobs. Unlike most of Europe (except maybe the UK), the US has been open to immigrants for centuries.

I think there is a selection effect here. If you are low income and want cushy social services, move to Sweden or NY or California. If you don't want to pay taxes to support these services, move to Florida or Nevada or New Europe.

If you want to work for a big company, move to Germany or France. If you want to start the next big company, move to Texas or North Carolina or (if they avoid the looming tax explosion) California.

I'd want to know where the most productive people want to live. Not the wealthiest or highest paid, but the ones who add the most value, create the most jobs and economic growth. Industrial research scientists or entrepreneurs would seem to fit this category.

PS: If you want to learn about economic freedom, try Chile. The professional class there seems to understand the importance of freedom, having seen the alternative.

scipio writes:

@ Joel

Your view on Scandinavia is too limited. Yes our economical freedom is smaller than in the US or, say, Chile. A quite significant part of our BNP is used by the state. On the other hand, we also have a very competitive private sector and a high individual freedom.

The way the Scandinavian states work enjoy a very big concensus in the population, say more than 80 % of the population do not want to change it significantly.

The way I see it is that the Scandinavian "way" works well for us, but we have a very special history and our approach may certainly not work well in other countries. Hence the critic by Bryan Caplan of the HDI is very relevant!

Regarding efficiency, the Scandinavians are actually quite good. Yes, we have a smaller BNP pr. inhabitant, but we also work significantly fewer hours: 37.5 hours pr. week and 6 weeks of holiday pr. year.

jd writes:

[Comment removed for supplying false email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Comments for this entry have been closed
Return to top