Inequality and Growth
Even if I have an otherwise crappy day, when I come home and there's a new book from Amazon or a new economic journal sitting in my mailbox, I brighten up considerably. The most recent arrival has been "Inequality and Growth, Theory and Implications" edited by T.S. Eicher and S.J. Turnovsky. It's a good read and highly recommended. Here I'll summarize and discuss the first article in the volume:
Francois Bourguignon – Growth Elasticity of Poverty Reduction (last link to the WB version of paper).
The issue is of course the old one - should development focus on growth, or on redistribution? Which one can do more to reduce abject poverty? Well, in order to answer that question we need to have some estimates as to how both growth and inequality affect poverty.
So first, a description of the paper and then some wanton speculation on my part. Just to be clear, here we're talking absolute poverty defined as living on less then 1$/day.
General results from literature – 1% increase in average income reduces absolute poverty by 3% (in World Development Report that elasticity closer to 2). But there is lots of heterogeneity across countries.
Basically this paper decomposes changes in absolute poverty into changes in mean income and changes in relative poverty (inequality) and examines how this elasticity varies by inequality levels and poverty levels (there's some problems with this approach which I might address in later posts).
Here's a table with some examples and summary numbers:
Some caveats – the Gini measure of inequality is not good at comparing economies of different sizes. Specifically under many circumstances, it tends to exaggerate inequality in large countries, like US for example. So in the above list, India probably has a lower level of inequality then the Gini would indicate (which means it’s a very equal country. Very equal and (still) poor country).
As a result if there’s a negative or a reverse-U shape between inequality and growth, in the very inegalitarian countries (Latin America, parts of SS Africa) redistribution is Bonus. For one, it makes for a more equal country, but it is also likely to increase growth. As inequality increases the elasticity of poverty reduction due to growth increases and the potential gains from redistribution (in terms of poverty reduction) are amplified.
________________________________________________________________
summary of the article ends and wanton speculation on my part begins here
________________________________________________________________
What does this mean?
Land.
More specifically land reform, as a prerequisite for development. Acemoglu and Robinson also note the role of unequal land ownership (and hence general unequal distribution of incomes as found in say Latin America) in preventing the transition to democracy and fostering a country’s propensity for revolutions, coups and other social upheavals (which themselves tend to have a negative impact on growth).
Basically, there’s “good” inequality and there’s “bad” inequality, and if you are a person who cares only about poverty/growth but not inequality per se (and to a first order approximation I’m just such a person) then you should worry about the latter but not the former. And the latter is very close tied to land and natural resources.
If you think about which countries are very unequal today, which have had series of revolutions and coups in their history and for which the most plausible case can be made for inequality acting as drag on growth you get the following list:
Most of Latin America, excluding Columbia. Mexico and Argentina for example both enjoyed fairly high economic growth during 19th century, a lot of it financed by European capital. But Mexico ended with a destructive Civil War/Revolution and subsequently a nondemocratic, somewhat populist government. Argentina’s growth ended when foreign capital flows dried up in the wake of WWI and the collapse of “The First Era of Globalization” during the interwar years and the emergence of the populist Peron, alternating with military coups in the aftermath.
Post WWII Mexico saw very uneven growth with periods of improvements alternating with periods of stagnation and occasional regress. Also, for all its populist rhetoric, the Mexican government did not alter the unequal distribution of income very much.
(download this dynamic chart from Sala-i-Martin’s webpage and look at the “three-peaked” distribution of income as late as 1970. It’s only since the early 80’s (and it should be noted, the initiation of pro-market reforms) that Mexico has become more equal though it still remains a “double peaked” country (so is US for that matter, it’s just that the high peak in Mexico is the poor one, whereas it’s the rich one for the US))).
(Yes, parentheses are my favorite punctuation symbol)
Argentina likewise saw stagnation and crises post WWII so that it moved from being ranked in the top 5 richest countries in the late 1800’s to being almost a “middle income” country today (it’s still however one of the richest economies in South America, though Chile’s been catching up).
Both countries basically never fully solved the problem of land reform and the inequality that results from it. In Mexico of course land reform was the main driving force behind Zapata, and because Carranza needed to consolidate his gains by co-opting a portion of rich landowners, they had him killed. Obregon continued Carranza’s land policies. As a result meaningful land reform never got under way. I know somewhat less about the economic history of Brazil but I’m pretty sure it fits the pattern.
Russia: Serfs didn’t get emancipated until second half of 19th century and even then on very unfavorable terms. Land reform and the plight of the peasants was a recurring theme in revolutionary movements throughout late 19th century and early 20th. A good part of the reason for the fall of the liberal democratic Kerensky government in 1917 (aside from its unwillingness to exit WWI) was its inability to carry out land reforms (this parallels the fate of Madero in the Mexican Revolution). While the Bolsheviks initially did make some “meaningful” moves in really redistributing land, the state collectivization of agriculture pretty much ended that resulting in the Great Famine, designed by Stalin as a means of repressing nationalist Ukrainians and stubborn independent mind peasants who had the nerve to expect actual gains from the revolution. After that Russia became the Soviet Union, a communist economy for which the regular rules don’t apply, so the example breaks down at this point.
Sub Saharan Africa: Here of course the key is colonialism. Whatever distribution existed before the Europeans got involved it got trumped by what came afterwards. There’s hardly a need to argue that the distribution of land in most of SS Africa was unequal as that was the very nature of colonialism. While the Western European countries became more democratic and equal back home, in SS Africa, it was Feudalism full throttle (which is why there’s nothing inconsistent about being a capitalist/libertarian anti-imperialist/colonialist. Some annoying lefties since Lenin have insisted that imperialism and capitalism were/are inexorably connected. Today they insist that trade between the poor countries and the rich ones is a form of neo-colonialism. Of course this is all bunk. The colonial system imposed on SSA countries was about as removed from capitalism as you can get without going SU style commie. In the post colonial period this did not change much. Power, and the distribution of land changed but the concentration of it did not. In some cases those who took over the land were former colonial administrators, either European derived, or a privileged native group. In others it was the revolutionaries who drove out the Europeans which got the spoils. In either case SS Africa remained a very unequal place, with at best “land reform” amounting to government ownership of land (Russian style collectivization). If you add in natural resources, of which SS Africa has plenty, into the definition of land the situation becomes even more exacerbated.
I should note here that I do not consider nationalization of land and natural resources by governments, like in SU, many African countries, or more recently in Venezuela to be “land reform”. This could be called “land reform” if the profits and spoils somehow made their way back to “the people” but, for many Public Choice reasons (all of them enumerations on the point “politicians steal like crazy because it is in their interest to do so”) they don’t. On the other hand in US "land reform" took a particularly nasty form.
Of course, in the end all of this is perfectly consistent with both classical and neo-classical economic theory, as well as standard Public Choice and Political Economy ideas. Land is (essentially, relatively) a fixed factor of production hence the ownership of land results in Ricardian rents. Political squabbles and fighting over the control of these rents results in both dissipation of resources (lower growth) but also usually ends up with a ‘winner takes all society’ with high inequality. Hence I would argue that it is not inequality per se that constraints growth and poverty reduction, rather it’s a third factor – the prominence of land/natural resources in the economy – which causes high inequality, low growth and the low elasticity of poverty reduction found in this paper.
There’s some problems with this paper, just as there are with pretty much any measure of ‘inequality’ (i.e. summarizing an entire distribution through a single number like with the Gini) to which I might return in later posts.
Somewhat relevant to this is also Evsey Domar’s The Causes of Slavery or Serfdom: A Hypothesis.(jstor)
By the way the guy with the coolest name ever was one of Zapata's generals.
Even if I have an otherwise crappy day, when I come home and there's a new book from Amazon or a new economic journal sitting in my mailbox, I brighten up considerably. The most recent arrival has been "Inequality and Growth, Theory and Implications" edited by T.S. Eicher and S.J. Turnovsky. It's a good read and highly recommended. Here I'll summarize and discuss the first article in the volume:
Francois Bourguignon – Growth Elasticity of Poverty Reduction (last link to the WB version of paper).
The issue is of course the old one - should development focus on growth, or on redistribution? Which one can do more to reduce abject poverty? Well, in order to answer that question we need to have some estimates as to how both growth and inequality affect poverty.
So first, a description of the paper and then some wanton speculation on my part. Just to be clear, here we're talking absolute poverty defined as living on less then 1$/day.
General results from literature – 1% increase in average income reduces absolute poverty by 3% (in World Development Report that elasticity closer to 2). But there is lots of heterogeneity across countries.
Basically this paper decomposes changes in absolute poverty into changes in mean income and changes in relative poverty (inequality) and examines how this elasticity varies by inequality levels and poverty levels (there's some problems with this approach which I might address in later posts).
Here's a table with some examples and summary numbers:
Some caveats – the Gini measure of inequality is not good at comparing economies of different sizes. Specifically under many circumstances, it tends to exaggerate inequality in large countries, like US for example. So in the above list, India probably has a lower level of inequality then the Gini would indicate (which means it’s a very equal country. Very equal and (still) poor country).
As a result if there’s a negative or a reverse-U shape between inequality and growth, in the very inegalitarian countries (Latin America, parts of SS Africa) redistribution is Bonus. For one, it makes for a more equal country, but it is also likely to increase growth. As inequality increases the elasticity of poverty reduction due to growth increases and the potential gains from redistribution (in terms of poverty reduction) are amplified.
________________________________________________________________
summary of the article ends and wanton speculation on my part begins here
________________________________________________________________
What does this mean?
Land.
More specifically land reform, as a prerequisite for development. Acemoglu and Robinson also note the role of unequal land ownership (and hence general unequal distribution of incomes as found in say Latin America) in preventing the transition to democracy and fostering a country’s propensity for revolutions, coups and other social upheavals (which themselves tend to have a negative impact on growth).
Basically, there’s “good” inequality and there’s “bad” inequality, and if you are a person who cares only about poverty/growth but not inequality per se (and to a first order approximation I’m just such a person) then you should worry about the latter but not the former. And the latter is very close tied to land and natural resources.
If you think about which countries are very unequal today, which have had series of revolutions and coups in their history and for which the most plausible case can be made for inequality acting as drag on growth you get the following list:
Most of Latin America, excluding Columbia. Mexico and Argentina for example both enjoyed fairly high economic growth during 19th century, a lot of it financed by European capital. But Mexico ended with a destructive Civil War/Revolution and subsequently a nondemocratic, somewhat populist government. Argentina’s growth ended when foreign capital flows dried up in the wake of WWI and the collapse of “The First Era of Globalization” during the interwar years and the emergence of the populist Peron, alternating with military coups in the aftermath.
Post WWII Mexico saw very uneven growth with periods of improvements alternating with periods of stagnation and occasional regress. Also, for all its populist rhetoric, the Mexican government did not alter the unequal distribution of income very much.
(download this dynamic chart from Sala-i-Martin’s webpage and look at the “three-peaked” distribution of income as late as 1970. It’s only since the early 80’s (and it should be noted, the initiation of pro-market reforms) that Mexico has become more equal though it still remains a “double peaked” country (so is US for that matter, it’s just that the high peak in Mexico is the poor one, whereas it’s the rich one for the US))).
(Yes, parentheses are my favorite punctuation symbol)
Argentina likewise saw stagnation and crises post WWII so that it moved from being ranked in the top 5 richest countries in the late 1800’s to being almost a “middle income” country today (it’s still however one of the richest economies in South America, though Chile’s been catching up).
Both countries basically never fully solved the problem of land reform and the inequality that results from it. In Mexico of course land reform was the main driving force behind Zapata, and because Carranza needed to consolidate his gains by co-opting a portion of rich landowners, they had him killed. Obregon continued Carranza’s land policies. As a result meaningful land reform never got under way. I know somewhat less about the economic history of Brazil but I’m pretty sure it fits the pattern.
Russia: Serfs didn’t get emancipated until second half of 19th century and even then on very unfavorable terms. Land reform and the plight of the peasants was a recurring theme in revolutionary movements throughout late 19th century and early 20th. A good part of the reason for the fall of the liberal democratic Kerensky government in 1917 (aside from its unwillingness to exit WWI) was its inability to carry out land reforms (this parallels the fate of Madero in the Mexican Revolution). While the Bolsheviks initially did make some “meaningful” moves in really redistributing land, the state collectivization of agriculture pretty much ended that resulting in the Great Famine, designed by Stalin as a means of repressing nationalist Ukrainians and stubborn independent mind peasants who had the nerve to expect actual gains from the revolution. After that Russia became the Soviet Union, a communist economy for which the regular rules don’t apply, so the example breaks down at this point.
Sub Saharan Africa: Here of course the key is colonialism. Whatever distribution existed before the Europeans got involved it got trumped by what came afterwards. There’s hardly a need to argue that the distribution of land in most of SS Africa was unequal as that was the very nature of colonialism. While the Western European countries became more democratic and equal back home, in SS Africa, it was Feudalism full throttle (which is why there’s nothing inconsistent about being a capitalist/libertarian anti-imperialist/colonialist. Some annoying lefties since Lenin have insisted that imperialism and capitalism were/are inexorably connected. Today they insist that trade between the poor countries and the rich ones is a form of neo-colonialism. Of course this is all bunk. The colonial system imposed on SSA countries was about as removed from capitalism as you can get without going SU style commie. In the post colonial period this did not change much. Power, and the distribution of land changed but the concentration of it did not. In some cases those who took over the land were former colonial administrators, either European derived, or a privileged native group. In others it was the revolutionaries who drove out the Europeans which got the spoils. In either case SS Africa remained a very unequal place, with at best “land reform” amounting to government ownership of land (Russian style collectivization). If you add in natural resources, of which SS Africa has plenty, into the definition of land the situation becomes even more exacerbated.
I should note here that I do not consider nationalization of land and natural resources by governments, like in SU, many African countries, or more recently in Venezuela to be “land reform”. This could be called “land reform” if the profits and spoils somehow made their way back to “the people” but, for many Public Choice reasons (all of them enumerations on the point “politicians steal like crazy because it is in their interest to do so”) they don’t. On the other hand in US "land reform" took a particularly nasty form.
Of course, in the end all of this is perfectly consistent with both classical and neo-classical economic theory, as well as standard Public Choice and Political Economy ideas. Land is (essentially, relatively) a fixed factor of production hence the ownership of land results in Ricardian rents. Political squabbles and fighting over the control of these rents results in both dissipation of resources (lower growth) but also usually ends up with a ‘winner takes all society’ with high inequality. Hence I would argue that it is not inequality per se that constraints growth and poverty reduction, rather it’s a third factor – the prominence of land/natural resources in the economy – which causes high inequality, low growth and the low elasticity of poverty reduction found in this paper.
There’s some problems with this paper, just as there are with pretty much any measure of ‘inequality’ (i.e. summarizing an entire distribution through a single number like with the Gini) to which I might return in later posts.
Somewhat relevant to this is also Evsey Domar’s The Causes of Slavery or Serfdom: A Hypothesis.(jstor)
By the way the guy with the coolest name ever was one of Zapata's generals.
posted by YouNotSneaky! at 5:56 PM
3 Comments:
Wow! This is A LOT of stuff to cover! No surprise then that it was gestating for so long, ~ 2 weeks since your last post.
I don't trust these cross-country estimates. "Econometric policy evaluation" and all. You can mention all of these notions and stories and mechanisms but in the end the estimate doesn't tell me much. It's a bad idea, I think, to try and use it for out of sample prediction or for actual policy.
One way that would give me some confidence would be to parametrize the steady state of a stochastic growth model with the Gini coefficient and see how the dynamics change. Of course, the challenge is to be able to introduce meaningful heterogeneity in a such a model. :-(
But even so, I have a feeling that much of the problems in non-G7 countries comes from imperfect financial markets, liquidity issues, lack of property rights, etc., things that you can't capture in a neat stochastic growth model.
Regarding growth... and maybe this is off-topic here... the Soviet Union also grew quite vigorously, in its own way. -- There's a challenge in aligning growth with welfare. You can get higher output but does that output satisfy actual preferences? (Consumption-wise, intertemporally, etc.) In mostly-free-market countries this is a non-issue, but in places where "strategic" planning is in place, I'm concerned.
Regarding your comments on land... How much do you sympathize with Henry George's stuff?
Now, to read all the links... See you again in 5-6 days. :-)
Yeah I wrote most of this a week ago in some way but then real work and life got in a way.
As far as the "econometric policy evaluation" I sort of agree. On one hand I think it's important to get these kinds of estimates. On the other it's also important to understand why inequality is linked to growth, which is where all the stuff about land came from.
Meaningful heterogeneity would be the key and the challange. Probably something like an OLG model (since Ramsey though much simpler is not very good at handling heterogeneity). Of course the problem with OLG models is that pretty much anything could happen.
The SU is the classic example of oversaving and the case that higher growth does not necesserily equal higher welfare, even ignoring political and distributional issues. A more contrarian position would be that the east Asian economies and Japan grew "too fast" since 1950.
And yeah, I'm sympathetic to Henry George though I think some people take it way too far. There used to be a commentator on Brad Delong who no matter what the topic was, always dragged poor Henry George into it. The solution for all the problems in the world was to tax land.
WOW GOLD, nice blog. its worth reading. I liked it very much
Post a Comment
<< Home