about me
game theory
... more
Subscribe Weblog

A couple of hours ago, Abiola Lapite (Foreign Dispatches), a rootless cosmopolitan with a libertarian weltanschauung, wrote
The post on Growth Theory and "Stylized Facts" at Mahalanobis also makes for interesting reading, not least because it ties into the work of several economists whose work I very much admire. As far as I'm concerned, there is no question in economics more important than how to get poor countries to grow faster and in a more sustained fashion. (boldface added)
I can't agree more with that! One of my favorite quotes comes from Robert E. Lucas who wrote back in 1988[1]:
"Is there some action a government of India could take that would lead the Indian Economy to grow like Indonesia’s or Egypt’s? If so, what exactly? If not, what is it about the “nature of India” that makes it so? The consequences for human welfare involved in questions like these are simply staggering: Once one starts to think about them it is hard to think of anything else." (emphasis mine)
The short version of this statement, which shows up in many lecture notes on economic growth, is "Once one starts to think about economic growth it is hard to think of anything else". (Tyler Cowen recently came up with an even cooler version of this statement: Once you start to think about economic growth it is hard I to be on the political left ;-D).

One obvious question that has attracted considerable attention among growth theorists was whether poor countries can eventually catch up to rich countries in terms of their overall standard of living and whether this is actually happening. As you can imagine, two groups of economists have emerged from this (still raging) debate. As Phillips and Sul[2] put it:
"[0]ne group belives it has found a black hat of convergence (evidence for growth convergence) in the dark room of economic growth, even though the hat may not exists (the task may be futile). A second group believes it has found a black coat of divergence (evidence against growth convergence) even though this object also may not exist (empirical reality, including the nature of growth divergence, is even more complex than the models used to characterize it)."
The figure given above shows five groupings of cross sectional averages of per capita real income for 120 countries (from the Penn World Table) over 1960 to 1989. The groups are selected according to initial income level, each panel refers to the same period, viz. 1960 to 1989.

Calculating the (continuous) growth rates yields*:

Poorest: 1.5 %, Poor: 1.2 %
Middle: 2.8 %, Rich: 3.1 %
Highest: 2.7 %

But, before getting excited, look at those figures:

The second figure shows the 2.5%, 50% and 97.5% quantiles of the bootstrap distribution (5000 replications) of the real per capita income for the same country groupings and the third figure shows the cross section average trajectories against the maximum and the minimum, using actual data.
These curves give some idea of the variability in the actual growth trajectories over time within these groupings. The outcomes indicate that some members of each group have substantial prospects of transitioning into higher growth groups over the 30 year period.
You can probably imagine how much sense standard econometric tools make if a couple of parameters (speed of convergence, technological progress) should actually be allowed to vary over time and across units. The figure given below shows the case of transitional divergence and ultimate convergenceIn. Economy 1 has a similar initial endowment as economy 2, but this country’s growth diverges initially from both economy 2 and 3.

{Beware: Studying economics can cause serious brain damage}

One of Peter Phillips' and Donggyu Sul's conclucions:
These findings corroborate some of the arguments made in Lucas (2002), where it is suggested that the world economy diverges during transition and then starts to converge. Lucas’s conjecture is based on heterogeneity in the speed of human capital accumulation. While it is not a sequential growth path but instead represents the average growth of clusters of countries with widely different incomes, our [first figure] is similar in shape to the 1,500 year growth path plotted by Lucas (2002). In the early stages of an economy’s growth, the cumulative human capital stock is low and growth is slow. As an economy learns, imports or creates technology, the economy grows faster. These stages of growth are reflected in Lucas’s growth path and in the [first figure].
related items:
Distance to Frontier - Big Deal!!!, Michael Stastny
Growth regressions and what textbooks don't tell you, Jonathan Temple

*Phillips and Sul came up with different growth rates; they probably calculated for each club the average real per capita income growth rate and not, as I did, the growth rate of the average real per capita income.

[1] On the mechanics of economic development, Robert E. Lucas, Journal of Monetary Economics, Volume 22, Issue 1 , July 1988, Pages 3-42
[2] The Elusive Empirical Shadow of Growth Convergence, Peter C.B. Phillips and Donggyu Sul, January 2003
Ostracised meinte am 7. Sep, 10:42:
is it possible that poorer countries have had statistically worse governance, even in relative terms, than richer countries? or would you/Lucas subsume that under 'human capital stock'. 
Mahalanobis antwortete am 7. Sep, 16:35:
would have to control for such factors as "governance" when running so-called "growth regressions/barro regressions".

At first the concept of "absolute convergence" came up, i.e. all countries converge to the same "steady state" (the steady state is reached when a country only grows at the constant rate of technological progress ("balanced growth path"), which was assumed to be exogeneous (determined outside the model). I guess Charles Jones, an economist, termed it "manna from heaven"). Then a couple of economist showed up and said something like: Hey, countries may have different structural parameters and they may converge to different "steady states" (for example, if one country has an optimal savings rate and another country has a suboptimal savings rate, then the country with the "dynamically inefficient" savings rate will never catch up to the country with the initially higher income. So, the concept of conditional convergence was born and economists included variables in their growth regressions to control for such differences when calculating the "convergence parameter". The next step was to include variables that proxy for the things you mentioned in your post. If "governance", "trade openness", "quality of institutions", "inflation rate", ... make a difference, then they have to be included in the model... otherwise the model would be misspecified and all those variables would show up in the error term... Actually I don't want to continue writing since I would remind myself that all those growth regressions are trash anyway. ;-D 
kimcil (guest) antwortete am 7. Apr, 12:38:
I liked Gladwell's distinction between insufficient facts and inadequate analysis, but calling the former a "puzzle" and the latter a "mystery" is pretty much backward from how people normally use the words. In a murder mystery, the detective has to dig up lots of clues. In puzzles, the facts are all in front of you normally, but you have to figure out the solution.
alfamart official partner merchandise fifa piala dunia brazil 2014
Unit Link Terbaik di Indonesia Commonwealth Life Investra Link 
maskodok meinte am 15. Jan, 01:21:
Interesting topic for a blog. I have been searching the Internet for fun and came upon your website. Fabulous post. Thanks a ton for sharing your knowledge! It is great to see that some people still put in an effort into managing their websites. I'll be sure to check back again real soon.
Mobil Sedan COrolla,Idrpoker.com agen Texas poker Online Indonesia Terpercaya, Mobil Sedan COrolla, Cipto Junaedy