Peter Henry on Growth, Development, and Policy
Jul 27 2009

Peter Blair Henry of Stanford University talks with EconTalk host Russ Roberts about economic development. Henry compares and contrasts the policy and growth experience of Barbados and Jamaica. Both became independent of England in the 1960s, so both inherited similar institutions. But each pursued different policies with very different results. Henry discusses the implications of this near-natural experiment for growth generally and the importance of macroeconomic policy for achieving prosperity. The conversation closes with a discussion of Henry's research on stock market reactions as a measure of policy's effectiveness.

RELATED EPISODE
William Easterly on Growth, Poverty, and Aid
William Easterly of NYU talks about why some nations escape poverty while others do not, why aid almost always fails to create growth, and what can realistically be done to help the poorest people in the world.
EXPLORE MORE
Related EPISODE
Angus Deaton on Inequality, Trade, and the Robin Hood Principle
Nobel Laureate in Economics Angus Deaton of Princeton University talks with EconTalk host Russ Roberts about the economics of trade and aid. Deaton wonders if economists should re-think the widely-held view that redistribution from rich nations to poor nations makes...
EXPLORE MORE
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.

READER COMMENTS

Tony G
Jul 29 2009 at 1:02am

Three comments:

1) In the second half of the podcast, the issue of growth rates came up and it was noted that the US typically grows at a 2 – 3% rate annually, whereas some countries achieve 7 or 10%. It seemed to me as if there was a “gee, why can’t the US get to that level?” As someone who studies Latin America, most of those 7 – 10% growth rates come in countries that are well below the production possibility frontier, specifically with lots of unemployed human capital. I would argue that when you have lots of unused resources laying around, it is fairly easy to get 7 – 10% growth, even over a sustained period of time. Growth is largely a function of getting unused resources in play. A growth rate of 2 – 3% would indicate that we are pretty close to the production possibility frontier and the only way to achieve greater growth is via technological innovation, which is a bit harder to achieve and doesn’t come in “10% spurts.” The countries with 7 – 10% growth show little innovation, but rather just borrow (or hijack) other countries’ innovations.

2. Early on the conversation set up a dichotomy between institutions and policy. Henry’s study is really nice in that it does look at a very good “natural experiment” that tends to isolate institutional effects — both nations start with British institutions at roughly the same time of decolonization. The focus then turns to policy. However, I fear that this podcast downplayed the role of institutions and I’m wondering if more is going on here. Institutional contexts can have a major impact on policy choices. As Russ noted, the policies pursued by the GOP and Democrats over the past 40 or so years have been relatively similar* and this is likely a function of our “first-past-the-post” voting system that forces a two-party system.

I would even argue that subtle institutional differences can have a major policy impact. And this is a rhetorical set-up for my question. Might it be that there were subtle institutional differences between Barbados and Jamaica that led Jamaica down a worse policy path? I truly don’t know. This is a question for others to answer if they do know. One thing that caught my attention early on dealt with the size of the country and post-slavery labor. In Barbados, slaves tended to become wage laborers, wherein Jamaica available land tended to create more of a homesteading environment. One would reason that wage rates would then rise in Jamaica, and if the government felt it needed to compete with other countries on wages, it would enact policies inimical to free labor. Could this have had an impact on the course of policy in these countries?

3. I found this podcast hard to follow. In part, it sounded as if Henry was talking more towards Russ than to a general audience. There tended to be a bit more “inside baseball” with references to specific papers and scholars. I like the podcasts with Munger and Boudreaux because even though there is a discussion between two economists, it seems as if both parties know they are talking to a broader audience. I didn’t get that sense here. Plus, there tended to be a lot more mumbling and “tangential footnotes” on the part of the guest. I found myself rewinding the podcast quite a bit.

* Footnote: While in comparative perspective I would agree with Russ that Republicans and Democrats have pursued similar policies, I would note that it has been Democrats who have largely been responsible for initiating major shifts in policy direction (FDR, LBJ, BHO?). Yes, Nixon followed a path of price controls, EPA, etc., but this seems more of an extension of Great Society policies. Interesting counterfactual: Had the Great Society not been implemented, would Nixon have introduced the EPA or price controls?

emerich
Jul 31 2009 at 11:10pm

A bit shocked at the dearth of comments. I guess to some extent the interview was “ho-hum” because we learned that within historically broad ranges policy doesn’t matter (tax rates, deficits etc.) but beyond those ranges–policy starts to matter! (Parenthetically, doesn’t it really matter only if policy cracks the top end of the range? What country suffered when it lowered income or corporate tax rates or tariffs or government spending or deficits “too far”?) In the sense that Henry defines the “Washington consensus,” is there serious doubt that deficits can get too big, or tax rates too high? Aside from among a few politically professional economist-doubters?

AHBritton
Aug 3 2009 at 7:11pm

One aspect that comes to mind when thinking about this idea of institutions vs. policy is the case of China. China has very poor protections of property rights, a command style economy, and not many institutional features that seem traditionally beneficial to a free and economically prosperous society. True there are more economic freedoms than their have been in the past, even so I’m interested to hear people’s thoughts on this situation. Why would limited economic reform in a command economy lead to such explosive growth? What is the relation between policy and institution in this situation? Does their command economy actually benefit them in this situation? I understand the argument that they had so much potential for growth that it doesn’t take much to cause growth. Does this mean China could have 30-40% annual GDP growth if they had made more radical changes? or even more? And if all this is the case does that mean their is an expected GDP a country should have? or is that too Keynesian?

I appologize if that got too broad… was somewhat train of thought and I figured I’d help out with the comments 🙂

Tony G
Aug 3 2009 at 11:11pm

In response to AHBritton:

1) Don’t think of the relationship between institutions and GDP growth as linear. Just because they have (somewhat) freer markets now and are averaging 10% annual growth doesn’t mean that if you had three times as many institutional protections of property rights that you would get 30% growth. In the short term (e.g., a year), you are constrained by upper limits on technology, labor efficiency, etc. As mentioned above, I know the Latin American cases better and I don’t recall any GDP growth rates topping more than 12% annually (and such rapid growth was generally followed by an equally steep recession). Anybody else think of countries topping 12% GDP growth?

2) I think in the case of China you are starting from a very low base of productivity, in large part politically induced (recall the Great Leap Forward and agricultural communes). Minor changes to rules on claimancy can have a dramatic impact on the incentives to be productive. While there aren’t strong institutions (e.g., court system) protecting property rights, there has been a greater expectation over the past two decades that if you make a profit you get to keep some portion of it (as compared to giving it all up to the state or commune). This started first in a couple enterprise zones in the mid-1980s and then expanded to the rest of the country in the past decade or so. That in itself can probably account for a big jump in the GDP.

3) While China is “wow-ing” the world with their rapid economic growth, I think much of it revolves around borrowing (or stealing) capital technology from other nations. Latin American economists refered to this as the “easy phase” of industrialization. Buy a bunch of steel plants or entice Nike to make footballs in China, combine it with labor that heretofore had been used pretty inefficiently in agriculture, and Voila! you have rapid economic growth. But there is something that is really missing in China’s growth story — innovation. Yes, by letting some people keep some profits and helping them acquire capital, you increase the incentive to work harder and smarter. However, if property rights are not too secure and the government is known to “steal” innovations, you have little incentive to take huge risks (and cocomitant costs) in coming up with something new. Name the last innovation to come out of China? You would be hard pressed to do so.

This is why I somewhat chuckle at people predicting that China will be the dominant economic power in the coming decade or so. I heard the same thing about Japan in the 1980s, and they tended to be more innovative than the Chinese are today. When your resources are employed about about 60% of its potential capacity in a society, it is easy to give the economy an amazing jolt by some easy institutional changes or basic infrastructural investments. But at some point in time, growth really relies on innovation and innovation, in turn, relies upon much more secure property rights than China has now.

Comments are closed.


DELVE DEEPER

About this week's guest:

About ideas and people mentioned in this podcast:Articles:

Podcasts and Blogs:


AUDIO TRANSCRIPT

 

Time
Podcast Episode Highlights
0:36Intro. [Recording date: July 15, 2009.] Recent years work trying to show the importance of institutions for a country's growth and its ability to escape from poverty. What do people mean by institutions and why are they important? Institutions: two classes. One: Legal institutions, closer to English common law system or French civil law? What's the difference? English common law gives strong protections to private property, creditors. French civil law, less emphasis on the protection of individual property. As economists, things that protect individual property rights gives individuals incentives. If I can be sure I will reap the benefits from the rent that land generates, I'll be more likely to invest--put structure on land, plant crops, etc. Laws which protect property rights in that way are more likely to generate economic activity. Also: other kinds of institutions; more specifically. Legal institutions literature, Andrei Schleifer; beyond to protections for creditors, specific set of laws, often determined by the nature of their colonial origins. Countries that were located in areas of the world that were more desirable for colonizers to settle in and live in--say, outside the malaria belt, Kenya vs. the Congo--were more desirable to live in if you were an Englishman or Frenchman in the 18th century. If you were likely to want to stay there, it was more likely that colonizers would try to put in place institutions that looked like their home institutions. What literature has done is to say, we've known since the work of Adam Smith that good institutions are correlated with good economic outcomes. Countries with good institutions have higher income than countries with bad institutions. Switzerland better off than Mozambique. But what causes what? Do favorable institutions lead to better economic conditions or is it the other way around. Or a third thing that causes both?
6:41So, what do we know or think we know about that relationship? Recent work tries to build on Adam Smith, Arthur Lewis: try to tease out the statistical causation. More recent work says: what do we believe? Believe if a country was colonized by the British, strong property rights, because outside the malaria zone, then no sense in which who you were colonized by 300 years ago can cause today's level of income. It could, but it couldn't go the other way: Today's income level couldn't be the cause of malaria 300 years ago. Article by Johnson and Robinson, shows that in fact there seems to be this causal impact of property rights. Shouldn't be controversial; hard to believe that not having the rule of law or having private property be of uncertain ownership could possibly be good for growth; has to be a question of magnitude. Strong consensus that private property is good for investment, improvement, but how large a factor? In statistical terms, what's the r-squared? There's a lot of variation in growth rates; how much of that variation might be explained by these institutional factors? Very controversial. Typically say if I go from a level akin to Zimbabwe's to Switzerland's, a two-standard deviation change, what is the effect on income? Have to believe you have controlled for everything else; hard over 250 years, peoples' differences. Leads to an average of $2000/capita. Don't put a lot of stock in those point estimates; but we want to have a good sense about whether these things really matter. Criticism of this literature is that it's so long run in its view that it's hard to draw policy conclusions. Suppose you could get away from the issues: how much confidence in the magnitude? Good news: you've identified a causal effect. Bad news: can do nothing to change your colonial origin. Difficult unless you are in Star Trek. Good and bad news about exogeneity. In theory you could alter your institutions. From perspective of the agents trying to institute policies--the governments--the question is if policy x is instituted, what's the likely effect? For that, hard to point to these studies for guidance. Much more short-run questions. Need to have other kinds of studies--not that these aren't valuable, qualitatively.
13:13Governments do need to make decisions over shorter lengths of time. Need to think less about pure cross-sectional studies and think more about time series. Not: is it the case that countries with better institutions have better growth, but whether when a country's institutions change, they have different growth? What is the impact of that change, if any? Paper with Conrad Miller: case study comparing Barbados and Jamaica. Why those two countries? From previous literature, if you have the right beginnings and inherit the right kinds of institutions then things turn out better on average than if you inherit French civil law, in a malaria belt. Two countries in very similar parts of the world, in the malaria belt, both British colonies, in fact slave colonies, initially sugar plantations--very similar beginnings. At the time of independence, both inherit English common law, Westminster parliamentary democracy, strong constitutional protection of private property--three things that are key items on Santa Clause list leading to good economic outcomes. Jamaica became independent in 1962, Barbados in 1966. Both a little over 40 years ago. Very much alike historically. Barbados a little physically smaller. Both English-speaking, culturally similar. Slaves were brought there to cultivate sugar. Slavery ended long before their independence, roughly 1835; long period of time post-slavery before independence; in that time, agriculture sugar-based, some diversification into other crops. Significant difference relates to the size. Courtney Blackman, as Governor of the Central Bank of Barbados, wrote about this difference: Barbados being much smaller, when the slaves were freed there, there wasn't much land for them to inherit or occupy, so they ended up working as wage earners for their former owners. In Jamaica, lots of uninhabited land. Jamaica is roughly the size of Rhode Island; Barbados roughly a tenth of the size. Consequence: former slaves going off into very rural areas in Jamaica. Blackman hypothesis: former slaves in Barbados became educated relatively more quickly than former slaves in Jamaica, became assimilated by proximity.
18:47Come independence, early to mid-1960s, what were their relative standards of living? Slightly higher in Barbados, per capita income something like $3700 in Barbados, $2600 in Jamaica; both relatively but not desperately poor. Over next 40 years, big income gap develops; by 2002, Barbados close to $10,000 per capita, Jamaica still in mid-$3000s. Blackman hypothesis is more human capital in Barbados; but not big enough differences in human capital. What's your hypothesis? Look at the policies as opposed to the institutions. Institutions, three things: constitutional protection of private property, English common law, Westminster parliamentary democracy--things you inherited. Policy: Macroeconomic policy choices made within that institutional framework--large or small fiscal deficit, fixed or flexible exchange rate, tax policy, free trade. Hypothesis: policies that the Jamaican government adopted, particularly in the early 1970s, that were quite different from the policies adopted by Barbados that led to the divergence. Wouldn't think that would be controversial, but it's been challenging for economists to show that good policy leads to good outcomes; and for it to lead to growth over sustained periods of time is even harder. Paper interesting because this is something close to a controlled experiment. Paul Collier podcast, statistical work prone to these statistical ambiguities. What findings? Frame more: One reason economists have focused on long-run growth rates since work of Solow--in which savings rates don't have a permanent on growth rates but only on levels. Versus like Paul Romer, finding things that affect growth rates as opposed to just the levels. Lesson of paper points out something Solow tried to underscore: even things that don't permanently affect growth rates can still have an enormous effect on standards on living; if you temporarily grow faster--for 20 years--can by compounding have a big effect. In Jamaica, oil price shock, 1973; in response, Jamaica decides to run bigger deficits, finance shock rather than adjust to it; borrow lots of money, capital markets don't spend it wisely; nationalize some industry; go on an unabashed policy of socialist expenditure in spite of worsening economic climate. Barbados is hit with the same shock; smaller island, imports less oil but not necessarily less per capita. Have their own adjustment problems; run big deficit for a year but retrench. Never actually in the midst of external current account deficits that led to temporary depletion of reserves, never imposed restrictions on trade or extreme restrictions on the use of foreign exchange. Jamaica, as response to foreign exchange shortage, put very strong restrictions on imports, spiraling list of import restrictions, rent-seeking and corruption. Jamaica should be rich--protectionism, all those jobs, and stimulus, should be booming. Do they boom a little bit? For a couple of years, driven by booming commodity prices coming out of the 1960s; but starting in 1973, goes into a 14-year economic decline. Barbados doesn't boom; grows by 1.2% per year per capita; but Jamaica is contracting by 2.3%, so a 3.5 percentage point difference for 14 years. All those policy choices that Jamaica makes are things that if you look at in context of Solow model, they are going to be temporary effects; but cumulative effect of those temporary effects over 10 or 20 years is big. Coming out, even if you both grow at 3% per year, you still start out very far behind. Presumes there is a natural rate of growth that if you had it right you could grow around 3%; can't grow at 7-10% a year. But if you blundered, you wouldn't be able to catch up. There are countries growing at faster rates than 3%, starting in 2007; but Jamaica wasn't one of them.
28:57For 14 years, Jamaica declining--did they change their policies at the end of fourteen years? Government gets booted out democratically in 1980. Some violence, but election itself free and fair. Michael Manley replaced by Edward Seaga, Jamaican Labour Party. More market-oriented; but didn't completely embrace that; not till Manley comes back in 1988-1989 and embraces the Washington consensus reluctantly--balanced budget, low tax rates, low inflation, open borders. Big thing Jamaica attempted to do was to start to stabilize the deficits and inflation; costs to make that transition. Jamaica very successful in getting the macroeconomic environment under control; inflation down to 10%, starting to run budget surpluses, starting in the late 1980s, taking till mid- to late-1990s. Bringing inflation down when you've lost credibility is hard. Stabilizing the macro environment is a necessary but not sufficient condition. Lots of micro issues not yet tackled. Path back out of decline doesn't necessarily have the same steepness as going in. If you have negative growth for 14 years and a population of relatively young men, guess what happens to crime rates. Behavioral, cultural challenges. Decline through 1987; after 1987 trying to get their house in order; slow positive growth.
33:00Barbados stable but not spectacular story of growth. Can get you a long way. Really see the difference when you go there. Henry from Jamaica originally, biased toward saying good things about country of birth. Barbados, high literacy rate, spending on health, education, etc. Barbados followed closer to the Washington consensus--should say it differently, it's an easy catch-phrase but lose half your audience, ideological bias or tool of aid agencies whose goal is to turn the world into crony capitalism. Just mean good policies: stable macroeconomic environment, low inflation, manageable deficits, incentive to invest. Entrepreneurship differences between the two countries? Role of uncertainty, rule of law: if you don't have it, incentive to invest in the future is lower. Example: Jamaica is not Zimbabwe, not a government that in the 1970s expropriated everything or threw everybody in jail. Jamaica had the rule of law. Incentives: in the 1970s, Jamaica did a levy on bauxite production--bauxite used in production of aluminum. Manley government felt they weren't getting a fair share of the royalties and decided they wanted to increase the tax rate. Bauxite companies took government to court; International Center for the Settlement of Disputes (ICSD) ruled that the government had not proceeded in a legal fashion. But Jamaican bauxite production never recovered to the level that existed before the levy. Are there still opportunities for bauxite? Yes; but fairly widely accepted that it created uncertainty--government can try to change the rules arbitrarily, so what else might they do? Effects elsewhere? Henry a U.S. citizen now, came in 1978; lots of skilled people left Jamaica during that time; father a chemist, found it hard to get raw materials and supplies; things more difficult to do. Manley: "Five flights a day to Miami if you don't like our policies;" and people got on. Stopping the flights is one way to stop the flight of talent. Pretty free, vocal press; not Zimbabwe.
39:58What lessons drawn from these stories? Policy really matters. That's not to say that institutions don't matter, but governments that are thinking about what to do to foster better economic outcomes should think about that macroeconomic policy really matters. Other lesson: path out of decline is much steeper--meaning it's harder to get out. You might think it would be the other way around. In terms of rebuilding society, very hard to undo those effects. Challenge first claim of policy mattering: Leamer podcast: last 50 or so years of American economic history is 2-3%, year in and year out; yes, some bad post-war recessions, but mostly growth sustained through all kinds of macroeconomic policy; big and low deficits, tax rate changes, low and high inflations, not much change in trade policy; but through all these changes, not much difference in outcomes. How does that reconcile? Perspective relative to the United States only. Democratic versus Republican regimes: policy variation much less than the rhetoric. Think about United States versus other countries--there you do see very big differences. Jamaica--in U.S. we take 2 or 3% growth for granted; but compared to other countries around the world, Latin America, Africa; policy does matter quite a bit. Generally true that there hasn't been that much variation between Democratic and Republican regimes in the United States, but you could argue that, say, top marginal tax rates--big debate in the 1980s, supply siders; came down quickly under Reagan; followed by Clinton administration that raised marginal tax rates--is claim that it was symbolic? Yes. Within certain ranges, these changes don't make that much difference.
45:20Before started taping, talking about all these different empirical and case study approaches--very complex system. Different policy regimes, different institutional regimes, different national characteristics of the native population. Trying to either cross-sectional or time-series growth differences--at a point in time across countries or over time within countries--is possible but seems to be a correlation between ideology of the researcher and the empirical. Too cynical or fair assessment? People are prone to confirmatory bias. Personal experience: if you think about the question of economic reform, roughly 25 years ago, Washington consensus; James Baker; IMF meetings. If you ask two economists the question if economic reform helpful or detrimental, get different answers depending on the ideologies; each can dig up a sophisticated study proving why right. More public discussion, public intellectuals, are less tethered by systematic evidence and more biased by personal views than journal publications. At that level, it does have an effect on policy. To make progress on this issue, need a more objective measure of whether policies are expected to help or hurt. Perspective: that's the key question; two ways to answer that: by debating, or debating with numbers and sophisticated statistical evidence--which can often be made to agree with view taken in advance. Trying to find a third way. No, on the empirical side, present evidence as honestly as able.
50:02Best unbiased barometer is to look at how stock markets in these countries respond to the announcement of policy changes. Not the end of the policy discussion; value, equity. Debt relief--Jamaica became heavily indebted during the earlier policies. As college student, believed of course debt relief would be good for Jamaica. But looking at evidence changed his mind. Basic point: Jeff Sachs, Paul Krugman, Stuart Meyers: debt overhang in corporate finance literature inspired same discussion in developing countries. Does a company's or country's having too much debt discourage more investment? Sachs: macro models showing that possible. Other argument: negative effect of debt relief. Empirical question. Brady Plan: 1989 for Mexico; liberate debt, countries had to accept other policies. Did create value. Problem: stock markets in these countries; something different about the poorest countries from the Brady countries. No market for private investment. Brady countries: as you write down debt, there is a market for private investment. Symptom that debt relief is designed to help is not there for the poorest countries. When you write down the debt of a poor country--say it owes $3 billion in debt servicing; receives $4 billion in aid flows, so on net $1 billion. People think if you write off the debt, they will get $4 billion. Not the way it works. The lenders have balance sheets too; when the lender writes off the debt, they have to write it off; less to lend. Multilateral aid more effective than bilateral aid. Debt relief results in crowding out effect. Complaint about empirical work is really about sophisticated statistical modeling. It's not that empirical work or facts don't matter; it's that obsession with complex statistical modeling to tease out causality looks like an intellectual dead end. Two issues: in some of these countries, stock markets are not large. Second, Justin Fox podcast, skepticism about stock markets as information. Upcoming book. Basic point: only need to believe that stock markets respond. The prices don't have to be exactly right. Efficient markets: There was a lot of bad news last year! If stock market didn't tank, would be worried. Don't want to only look at one economic variable. Stock market is a way to evaluate what people will expect to happen. Does what people expect actually occur? Want to look at average. Researcher's job is to bring more clarity to what you need to believe in order to hold onto your pet view.