Quality institutions and growth are not so well-connected

by on April 6, 2013 at 6:25 am in Economics, History | Permalink

Here is a passage from a new paper from Lant Pritchett and Eric Werker:

One of the puzzles to be reconciled is that although “institutions” are associated with long run growth rates, their predictive power for short-to medium-run (5 to 10 years) growth, or for growth accelerations is very weak. (Khan, 2007) makes the distinction between “market supporting governance” versus “growth promoting governance.” He points out that, while “governance” measures are correlated with levels of GDP per capita, there is little or no predictive power of the current level of institutions for future growth rates.
We illustrate this by comparing the level of income and a measure of “governance” versus the growth rate and that same measure of governance and the growth rate and the change in the measure of governance. Whereas the first is strong the second is quite weak and the third near zero.
…Essentially all rich countries have reasonably high “quality of government” and all countries with high “quality of government” are rich.
…But among countries with the same governance there are massive differences in growth (e.g. China versus Cote d’Ivoire) and countries with rapid growth (above .6) the ‘quality of government’ ranking ranges from .15 (Indonesia) to .9 (Singapore).

…Figure 7c shows there is no link at all between the improvement in ‘quality of government’ and economic growth 1984 to 2004. A country like Uganda has massive improvement but exactly average growth, China has massive growth and no improvement at all, Malaysia saw QOG worsen but growth well above average, etc.

The relevant figures start at around p.30.  Was it Jeff Sachs who put it this way?: Go back to 1960 and try to measure the quality of institutions any way you wish, knowing of course in advance which countries end up doing well.  Can you find any measure at all which predicts subsequent growth?  This is a tough problem for we economists.

There is another sense in which institutions have to be the causal factor, in which case we are very far from having a sense of how to measure good institutions.

1 Hoover April 6, 2013 at 7:17 am

I believe I remember you once posting a study that correlated educational standards at the beginning of the twentieth century with wealth at the end. It mentioned a few Scandinavian countries, but I forget the details.

It seems obvious to me that institutions matter a lot, but that it takes a long time to develop a country. It took Britain two hundred years to change from not very developed to highly developed. Should we expect other countries to make the same journey in only twenty years?

2 Millian April 6, 2013 at 7:56 am

Why not thirty years? Lots of that British development was about generating new ideas which can easily be imported or copied nowadays, like internal combustion, and many East Asian countries grew rapidly in a golden generation.

3 prior_approval April 6, 2013 at 9:42 am

‘like internal combustion’

Are you sure about that? – because the top internal combustion motor designs that come to my mind are called Otto and Diesel.

4 ricardo April 6, 2013 at 7:43 pm

Like he said, easily imported or copied.

5 Hoover April 7, 2013 at 2:59 am

I take your point, but I’d argue that even “simple” technology like the internal combustion engine isn’t easily copied.

My starting point is that *I* couldn’t personally copy it easily. I don’t have iron ore and a steel mill, a re-roller and a cutter. I don’t have the measuring instruments needed, nor even the car to put it in, let alone the knowledge of how to use the equipment.

I think an internal combustion engine is *very* difficult to copy. Like a pencil would be. And I’d probably find it impossible to copy a microchip in twenty years.

Almost everyone in a less-developed country is in the same position of great ignorance, so it would be difficult to contract out the parts. Even if they’re not, they’ve been assigned to improving agriculture or health or some field it would be hard to transfer out of. Apply this to the hundreds of technologies a developed economy “needs” and it becomes clear why technological progress is so difficult to achieve in a short time.

Sure, trade means you can import this sort of thing, but to pay the bills you then need to make or grow or dig something up that other countries need.

6 prior_approval April 7, 2013 at 3:31 am

Or like he actually said – ‘Lots of that British development was about generating new ideas….’

The British did not generate either Ottö’s or Diesel’s designs.

7 TallDave April 6, 2013 at 10:14 pm

It’s worse than that, it took humanity thousands of years to arrive at a contingent moment in history where inclusive politoeconomic institituons could arise. And that’s one of those “weak anthropic principle” problems because only a world that had already seen societies achieve such a state would be able to look back and wonder how improbable it was (based on the fact only a handful of states have evolved them since 1900, it looks pretty unlikely).

It’s true some states like South Korea or Japan were able to rapidly modernize, but they had societies that were already receptive to the idea elites should do something than loot the country (and Japan’s pre-WW II government had no problem with looting, pillaging, and raping the rest of the Pacific Rim).

I was a lot more hopeful about the world’s future before I read Acemoglu. Few elites are interested in raising the living standards in their country.

8 Michael G Heller April 6, 2013 at 7:35 am

“We are very far from having a sense of how to measure good institutions” — Yes, that’s a significant issue. It may be irresolvable. But I don’t see why it’s an obstacle to continuing the historical analysis of which institutional features were causal, and how to speed them up.

9 Frederic Mari April 8, 2013 at 7:07 am


I think it’s really hard to say what institutions maximize growth and how. Let’s take China. It’s quite obviously very corrupt. OTOH, their mercantilism and huge internal market meant that, despite that corruption, they can generate eye popping growth.

Basically, I suspect that which institutions end up being useful vary a lot and in some cases, there’s the issue of ‘just so’ story-telling i,e Singapore develops thanks to its outstanding government and free markets, China because of its government drive to industrialise and export, Indonesia because… etc.

10 J April 6, 2013 at 7:40 am

Can you find any measure at all which predicts subsequent growth?

National mean IQ, e.g., [1], [2].

11 david April 6, 2013 at 7:51 am

Flynn. Plausible explanations for Flynn – increased education investment, increased nutrition, lowered disease load – are all themselves caused by growth. You would be observing that growth predicts growth.

12 ladderff April 6, 2013 at 11:52 am

Yeah, J, you bad person. Move along; these aren’t the droids you’re looking for.

13 prior_approval April 6, 2013 at 7:44 am

‘We are very far from having a sense of how to measure good institutions’

Sounds like a job for political economists – which strangely enough, all economists were, until the later part of the 19th century.

14 Millian April 6, 2013 at 7:57 am

There are probably more political economists today than there were in the mid-19th century, though, so that doesn’t explain the paucity of explanations.

15 prior_approval April 6, 2013 at 9:45 am

Care to name any to Prof. Cowen? – he seems to be searching for exactly that sort of person (the Virginia School of public choice is obviously absolutely inapplicable to this discussion in all but the most negative fashion).

16 steve April 6, 2013 at 8:14 am

I suspect its a case of institutions being to broad a term. Maybe some institutions matter a lot, and others not so much. So if you have most of them and score well on the academics list but are missing a key one then maybe it doesn’t correlate so well. I don’t claim to know which is which but if I had to guess it would be that some mix of low violence and getting to keep most of the gains from your work is the key. It wouldn’t have to even be the same institutions that accomplish this task in different countries. Maybe one country is low violence due to a solid uncorrupt police force and in another its due to a strong church and religious conviction.

17 Bill April 6, 2013 at 8:21 am

I thought you would have quoted this conclusion instead from the abstract of the paper:

“Whether economic elites use their influence activities with political and bureaucratic elites to create more possibilities for economic structural transformation or, conversely, use their power to entrench their privileged position will, to a significant extent, determine whether episodes of rapid growth can be sustained, will peter out, or even be reversed. The mechanisms for elite commitment to sustained inclusive growth are discussed.”

It’s economic elites…not institutions…and what they decide to do with those institutions is what mattered.

18 sam April 6, 2013 at 8:27 am

Following the Noahopinion post you linked to the other day, why not start by looking for institutions that are congruent for manufacturing exports.

19 Rahul April 6, 2013 at 8:45 am

>>> A country like Uganda has massive improvement but exactly average growth, China has massive growth and no improvement at all, <<<

How do they measure the quality of government? Sounds like a very subjective measure. Does ideological bias color this?

20 Jacob A. Geller April 6, 2013 at 3:29 pm

Or, for that matter, how do they measure GDP? We (economists) have not done a great job measuring GDP either, in a lot of the countries in, say, Figure 7c: http://www.amazon.com/Poor-Numbers-Development-Statistics-Political/dp/080147860X

So neither variable is being measured very accurately.

21 Rahul April 6, 2013 at 8:52 am

Go back to 1960 and try to measure the quality of institutions any way you wish, knowing of course in advance which countries end up doing well. Can you find any measure at all which predicts subsequent growth?

One big problem would be what time horizon to use. Countries that had low growth on a 1960-1980 window might have high growth on a 1960-2000 window. So no matter what set of predictors you used the prediction needs to be different at different points in time.

That’s a time series or maybe a Markov Chain where every state depends on the previous state. Unless growth were always a convergent phenomenon in the long run I don’t see how we can expect a few 1960-measured parameters to ever predict it.

22 Rahul April 6, 2013 at 9:08 am

The whole exercise looks silly to me. Take this snippet from the article’s abstract:

“Two key unanswered questions in theories of growth are (a) why some countries successfully initiate episodes of rapid growth while others suffer extended stagnation and (b) why some countries are able to sustain growth episodes over many decades of rapid (or steady) growth while other growth episodes end in reversion to stagnation or collapse. “

If you replaced “country” by “firm” could you do any better? Can one take 1960 parameters for listed firms, say, location, sector, number of employees and whatever else you want and then do a good job predicting which firms will have high and low growth over 50 year horizons?

If we cannot even do it for firms what makes us ambitious enough to attempt it for a nation. Or does somehow aggregation and the law of large numbers step in and help the modeler?

23 Seth April 6, 2013 at 9:43 am

I agree.

Though, I might suggest a couple of factors worth looking at. For example, how widely do trials and experiments happen and what are the feedbacks to success and failure?

24 Spencer April 6, 2013 at 11:08 am

Very few firms that are now in the Dow or the S&P 500 were publicly traded firms some 75 to 100 years ago.

25 Willitts April 6, 2013 at 12:01 pm

I began a comment using a child’s development based on his environment or a firm in its early stages. I ceased because I don’t think the analogies are apt. Economies consist in trillions of transactions, and development is a very complex interplay. Measures of growth and development are more macro level. So looking at an individual is a poor exemplar for an economy. That would be like examining the conditions affecting the health and well being of a single cell or organ in a body or an individual person and using it to explain the health and well being of a species (perhaps a different one).

I agree with your first criticism that the analysis is very sensitive to the time span of analysis. A Markov chain might be a suitable model. An economy has many exogenous and endogenous factors that affect growth. Also, there is a continued conflation of growth and development as sell as issues of sustainability.

26 Careless April 7, 2013 at 4:29 pm

Firms are a hell of a lot more competitive than countries. Just looking at American states, that,s a much easier task than corporations.

27 *dan April 6, 2013 at 9:44 am

Maybe the correlation is inverse, Maybe in the long-term prosperous countries with x amount of growth develop stable institutions.

28 Willitts April 6, 2013 at 12:05 pm

Or in both directions.

Patent law and a legal system might contribute greatly to innovation in early stages of development but become a hindrance to development in later stages.

Legal systems are notorious for their ratcheting effect because of stare decisis. This ratchet provides stability and predictability, but becomes ever tightening control beyond the intentions of each click of the ratchet.

29 *daniel April 8, 2013 at 3:12 pm

Do you have any evidence for this? I’m genuinely interested. I’d be fascinated to see some literature or examples of states that don’t have these things and then see whether innovation happens.

30 kerokasta April 6, 2013 at 11:23 am

I don’t understand that last sentence.

31 Bill April 6, 2013 at 11:48 am

I am amazed at the post and the comments.

The post implies that it is the institutions which fail, but if you read the abstract, what the paper says it is the ELITES that FAILED–some acted to promote growth, others to protect what they had.

Amazing. A Failure of ELITES cast as the failure of institutions they directed and controlled.

Maybe we should look at the economic elites.

32 ricardo April 6, 2013 at 7:45 pm

Perhaps Tyler thinks the behaviour of the elites is endogenous.

33 Willitts April 6, 2013 at 11:49 am

I believe it all comes down to property rights, and there is no reason to believe that an effective institutional framework will show immediate results. Just as Huntington demonstrated that democracy takes hold with fits and starts, institutions often suffer setbacks and reform themselves.

Looking at the causes of setbacks is more likely to reveal the conditions that are missing. For example, the Chilean recession after the first wave of privatization revealed that heavily concentrated and integrated industrial and financial firms increased risk and amplified business cycles.

34 Paul Holden April 6, 2013 at 1:08 pm

One of the major factors behind China’s growth take off was the introduction of property rights in land, albeit in a round about way. If that is not a huge improvement in institutions, I do not know what is

35 Steve Sailer April 6, 2013 at 1:13 pm

Tyler tautologizes:

“There is another sense in which institutions have to be the causal factor, in which case we are very far from having a sense of how to measure good institutions.”

Yes, but this paper doesn’t understand the full genius of Acemogluism, which is that anything bad that happens must the result of bad institutions, just as anything good that happens must be the result of good institutions:


36 Jacob AG April 6, 2013 at 3:35 pm

For all we know Africa’s current level of development and GDP growth rates are 75% determined by the fact that Westerners et al (Indians, other Africans, etc.) effectively kidnapped about half the population of Africa over the course of a few hundred years, a few hundred years ago.

Viewed that way, the fact (if it even is a fact) that today’s improvements in governance do not predict tomorrow’s growth rates, might be a myopic way of looking at whether/how institutions matter.

(I.e. in the long run we might be dead, but our grandchildren aren’t.)

37 TGGP April 6, 2013 at 4:21 pm

My understanding is that the western slave trade actually involved fewer numbers than the eastern slave trade, to the Ottoman empire. Of course, those were not profitable agricultural workers but castrated servants as a form of conspicuous consumption. I recall seeing a paper attempting to check the degree to which an African country was affected by the various slave trade routes with later measures of institutions or economic growth, but now I can’t recall the name of the paper, authors or what the conclusion was. I’d appreciate it if another commenter here linked to it.

38 Douglas Knight April 6, 2013 at 6:54 pm

Nunn, Nathan. “The Long-Term Effects of Africa’s Slave Trades.” Quarterly Journal of Economics, 123 (2008): 139-176.

The trans-atlantic slave trade was short in duration, but very large.

39 TallDave April 6, 2013 at 10:02 pm

It’s a very, very bad idea to try to evaluate a rate of change based on a static variable.

China’s growth rate is higher because its institutions used to kill 50 million people and such a basket case economically they tried to force the population to make steel in their homes (!), but are now merely banana-republic corrupt with a semblance of free markets. They’re still much poorer than Mexico, and they will remain so until their instutitons improve.

40 TallDave April 6, 2013 at 10:05 pm

Can you find any measure at all which predicts subsequent growth?

Yes, if you correct for PPP GDP per capita starting levels, the correlation to institutions is very strong. People tend to forget an economy’s absolute per capita production says far more about living standards than its rate of growth.

41 PG April 8, 2013 at 7:36 am

China had no improvement in institutions since 1984? This seems to say more about their measure of institutions than about China…

Comments on this entry are closed.

Previous post:

Next post: