Vietnam was just the median of these selected growth episodes- because Pritchett in his example uses quite a big growth episode. Pritchett calculates the NPV gain from growth acceleration per person from this median case as $6,914. This is for illustrative purposes, picking Vietnam has no special significance here. “To be affected by a think tank” also has no special significance, we didn’t check whether this growth episode was likely affected by a think tank.
These are selected by Pritchett:
“These are a list selected the largest episodes of growth acceleration. Source: Selected episodes. Author’s estimates from estimates in Pritchett, Sen, Kar, and Raihan 2016”
so… re your question:
When I look at the those figures in Appendix A, though, it seems like the median growth episode calculated using PRM (without reference to dollar size) is somewhere around Ecuador’s negative growth in 1978, which doesn’t seem like it would line up even with the conversion to $PPP.
Yes, this is likely largely due to Vietnam having a roughly ~10x higher population and being 10x poorer back then.
I think it is okay to use, as Pritchett does, these selected growth episodes, because if one wants to maximize effectiveness using policy one can strategically only look at big poor countries. One could further look at only those countries where growth is sluggish and perhaps where economic policy is particularly bad.
I write about this in the appendix:
Because effective altruism often tries to focus on the poorest countries, where a dollar goes 100x further than in rich countries, there is perhaps most hope for growth diagnostics.
So perhaps Duflo is right in that “Growth is likely to slow, at least in China and India, and there may be very little that anyone can do about it.” And this is actually born out in China’s and India’s performance on the World Bank’s Doing Business indicators, where they score 63th and 31st out of 189 countries, though being relatively poor. Thus, there seem no low hanging fruit to improve their economic policy.
However, below I show a table where I multiply population size of every country by their poverty multiplier (i.e. $1 is worth x times more going to this country than to the richest country in the sample. See appendix 2 of this doc for more info). This can then be ordered by the utility created by increasing GDP per capita by $1. India comes out on top because of its large population (1.3bn) and relatively low GDP per capita ($6,574). China comes 3rd, because though it has a large population, it is already relatively rich ($15,531). Recall that the problem is that we might not know how to increase growth in India and China.
However, there are many smaller very poor countries in the top 10 sample such as DRC and Ethiopia—very poor countries with 100 million population. This can then also multiplied further by neglectedness/tractability criteria. For instance, in a country’s ranking on the WB Doing Business ranking divided by GDP. There one can see that, relative to its GDP per capita, China already does quite well on the Doing Business ranking. However, the DRC and Ethiopia do poorly on the doing business ranking, even relative to their GDP. These countries could be most cost-effective for economic policy assistance.
Thanks for your response! I still have some confusion, but this is somewhat tangentially related. In your CBA, you use an NPV figure of $3752bn as the output gain from growth. This is apparently derived from India’s 1993 and 2002 growth episodes.
The CBA calculation calculates the EV of the GDP increase therefore as 0.5*0.1*3572 = $178.56 bn. You acknowledge elsewhere in your writeup that efforts to increase GDP entail some risk of harm (and likewise with the randomista approach) so my confusion lies with the elision of this possible harm from the EV calculation.
Even if the probability that a think tank induces a growth episode—e.g. the probability that a think tank influences economic policy in country X according to its own recommendations—is 10%, then there is still obviously a probability distribution over the possible influence that successfully implemented think tank recommendations would have. This should include possible harms and their attendant likelihoods, right?
I recognize that the $3,572bn figure comes directly from Pritchett as part of an assessment of the Indian experience, but it’s not obvious to me that the number encapsulates the range of possibilities for a successful (in the sense of being implemented) intervention. I may be missing something, but it seems to me that a (perhaps only slightly) more rigorous CBA would have to itself include an expected value of success that incorporates possible benefits and harms for both Growth and Randomista approaches in the line of your spreadsheet model reading “NPV (@ 5%) of output loss from growth deceleration relative to counter-factual growth.”
I understand that what you’re envisioning is a sort of high-confidence approach to growth advocacy: target only countries where improvements are mostly obvious, and then only with the most robustly accepted recommendations. I still think there is a risk of harm and that the CBA may not capture a meaningful qualitative difference between the growth and randomista approaches. In principle, at least, the use of localized, small-scale RCTs to test development programs before they are deployed avoids large-scale harm and (in my view) pushes the mass of the distribution of possible outcomes largely above 0. No such obstacle to large harms exists, or indeed is even possible, in the case of growth recommendations. Pro-growth recommendations by economists have not been uniformly productive in the past and (I think) are unlikely to be so in the future.
I still favor this approach you suggest but, given the state of the field of growth economics—and the failure of GDP/capita to capture many welfare-relevant variables that you cite at the end of the writeup—I’d be keen to see more highly quantified conversation around possible harms.
Excellent comment—strongly upvoted for engaging with the data.
The sheet where we calculated the median growth episode within the spreadsheet is here:
https://docs.google.com/spreadsheets/d/1VcQ2r5zuCztd1_2vRscK8UOEAiQqhvvhkJVfagCzpqQ/edit#gid=1331750623&range=D26
Source: Pritchett, Labor p23
Vietnam was just the median of these selected growth episodes- because Pritchett in his example uses quite a big growth episode. Pritchett calculates the NPV gain from growth acceleration per person from this median case as $6,914. This is for illustrative purposes, picking Vietnam has no special significance here. “To be affected by a think tank” also has no special significance, we didn’t check whether this growth episode was likely affected by a think tank.
These are selected by Pritchett:
so… re your question:
Yes, this is likely largely due to Vietnam having a roughly ~10x higher population and being 10x poorer back then.
I think it is okay to use, as Pritchett does, these selected growth episodes, because if one wants to maximize effectiveness using policy one can strategically only look at big poor countries. One could further look at only those countries where growth is sluggish and perhaps where economic policy is particularly bad.
I write about this in the appendix:
Thanks for your response! I still have some confusion, but this is somewhat tangentially related. In your CBA, you use an NPV figure of $3752bn as the output gain from growth. This is apparently derived from India’s 1993 and 2002 growth episodes.
The CBA calculation calculates the EV of the GDP increase therefore as 0.5*0.1*3572 = $178.56 bn. You acknowledge elsewhere in your writeup that efforts to increase GDP entail some risk of harm (and likewise with the randomista approach) so my confusion lies with the elision of this possible harm from the EV calculation.
Even if the probability that a think tank induces a growth episode—e.g. the probability that a think tank influences economic policy in country X according to its own recommendations—is 10%, then there is still obviously a probability distribution over the possible influence that successfully implemented think tank recommendations would have. This should include possible harms and their attendant likelihoods, right?
I recognize that the $3,572bn figure comes directly from Pritchett as part of an assessment of the Indian experience, but it’s not obvious to me that the number encapsulates the range of possibilities for a successful (in the sense of being implemented) intervention. I may be missing something, but it seems to me that a (perhaps only slightly) more rigorous CBA would have to itself include an expected value of success that incorporates possible benefits and harms for both Growth and Randomista approaches in the line of your spreadsheet model reading “NPV (@ 5%) of output loss from growth deceleration relative to counter-factual growth.”
I understand that what you’re envisioning is a sort of high-confidence approach to growth advocacy: target only countries where improvements are mostly obvious, and then only with the most robustly accepted recommendations. I still think there is a risk of harm and that the CBA may not capture a meaningful qualitative difference between the growth and randomista approaches. In principle, at least, the use of localized, small-scale RCTs to test development programs before they are deployed avoids large-scale harm and (in my view) pushes the mass of the distribution of possible outcomes largely above 0. No such obstacle to large harms exists, or indeed is even possible, in the case of growth recommendations. Pro-growth recommendations by economists have not been uniformly productive in the past and (I think) are unlikely to be so in the future.
I still favor this approach you suggest but, given the state of the field of growth economics—and the failure of GDP/capita to capture many welfare-relevant variables that you cite at the end of the writeup—I’d be keen to see more highly quantified conversation around possible harms.