Don’t we know that all that matters for reducing poverty is growth, especially after China? And therefore we development economists should focus on the things that make growth happen: Macro policy and creating the right institutional environment. And not bother with the micro evidence…
No, no, and, as the expression goes, no. Every step in that syllogism is wrong, and, I will argue … each step is probably more obviously wrong than the previous one.
It is rather ironic that just as the annual meeting of the ADB convened in Manila to discuss the “Asian century” emphasizing the role of the region in propping up the global economy that a survey by the Social Weather Station on self-reported poverty showed that a majority of Filipinos (55% in the first quarter of 2012 compared to 45% in the previous quarter) still considered themselves poor. Business leaders and policymakers seemed unflustered though as they expressed confidence in our long run prospects for growth and development.
After more than sixty years of working with both bilateral and multilateral aid agencies as well as non-governmental organizations, you would think that a good understanding of the links between growth and poverty reduction would exist. Unfortunately, there is strikingly very little we know based on the evidence that is available.
At the very least, all that we do know is that the so-called trade-off between growth and poverty reduction does not hold. From cross country analysis, we find that fast growing countries tend to see poverty decline. But that is about it. We do not know which causes which. Is it growth that reduces poverty or does poverty reduction provide the initial conditions for growth to occur?
To be clear, we don’t know if: (a) countries that grow faster, reduce their levels of poverty quicker (growth reduces poverty), (b) countries that reduce poverty subsequently grow faster (poverty reduction causes growth), or (c) growth without poverty reduction becomes unsustainable politically and therefore leads to cases of growth interrupted (both growth and poverty reduction have to occur simultaneously).
The previous government of the Philippines for instance spent the better part of the last decade focusing on growth through public spending and foreign investments. The incidence of poverty however did not decline which led to the pilot-testing of a conditional cash grants program to indigent households near the end of its second term of office. Its political difficulties stemmed from another source: perceptions of illegitimacy in the way it had acquired and held on to power.
The subsequent government has placed a greater emphasis on fixing perceived areas of corruption while rolling out the conditional cash grants program of its predecessor, but has found not only growth to have slowed, but (self-reported) poverty to have risen as well, at least at the onset.
This to-ing and fro-ing from one policy agenda to another demonstrates why it is so important to know the answer to the question how are growth and poverty related. Otherwise we could be wasting much of our time and resources promoting one type of reform when in fact we should be promoting another.
We know precious little about what promotes growth to begin with. It would be a shame investing so much in it (getting an investment grade credit rating for instance through fiscal austerity) when what we should be doing is throw everything we’ve got at something else.
It doesn’t help that what we thought we knew has been subsequently disproven or found to be flawed. For example, back in the 90s we were told that growth (or the lack of it) was mostly explained by adequate (or inadequate) levels of savings and investment in physical infrastructure, human capital and population. It turns out those factors only account for about a third of the growth. The rest remains unexplained.
The implications of this are that even if the government were able to take advantage of the fact that for once in our nation’s history we are no longer a net debtor, but a net creditor nation, and competently implement the roll-out of its public private partnerships, its education and health reforms and some variation of a reproductive health program, growth of the kind that it hopes for would still not be guaranteed.
If we were to compile all the findings of empirical work on cross-country growth rates, we would find that a total of 145 variables have been identified as significant, contributing either positively or negatively to growth. Given the number of countries to be observed and the range of variables to be tested, no satisfactorily robust method exists that would allow us to test all these variables all at once to measure their relative importance.
Even if we relied on partial correlations that tell us that policies, institutions and anti-corruption measures significantly affect growth, we still would not know which policies, institutions and forms of anti-corruption matter most. The data sets for testing such suppositions simply do not exist. Macro analysis simply is not viable. It exposes much policy work to the accusation that it is merely based on conjecture, ideology and politics.
Suppose we were to test proposed solutions at the micro level using micro experiments? Random evaluations using control and test groups might hold the key. Applying for instance new forms of administration across a range of local government units might help us rule out or rule in certain reforms. Evaluating specific policies like the Pantawid Pamilya as it rolls out might provide evidence as to how our version of conditional cash grants is either effective or ineffective in different settings throughout the country.
(Note the distinction I made between Pantawid Pamilya and other similar programs such as Brazil’s Bolsa Familia. Knowing that a version of conditional cash grants works in one setting does not necessarily prove it will work everywhere else. We should be cautious in accepting the claims made by Jeffrey Sachs, a speaker at the ADB meeting, who believes that we already know how to fight poverty in every setting throughout the world.)
When it comes to promoting growth through industrial policy (targeting poverty through industry and employment programs), we were told by the Washington Consensus not to experiment anymore. Growth came by lowering trade barriers and improving governance. They claimed that “one size fits all”. Countries had to craft their policies in a way that followed “best practice” which meant letting the free market allocate resources and determine our place in the value chain of global production.
What our experience and that of Latin America and Sub-Saharan Africa in adopting this so-called consensus points out, what we should have been pursuing was not the principle of best practice (our laws and regulations are often regarded as such), but the one of “best fit” as highlighted by the East Asian growth experience.
The countries in that region didn’t accept the Western capitalist mould of economic governance, but developed their own forms based on local institutions and tradition. This made the process of development more acceptable and sustainable. If we are to follow their path, then perhaps we should not be so hesitant to experiment as they did.
The problem is that our policy elite suffer from an identity crisis. They are habitually more Asian in their practices, but they seek to pass themselves off as Western. As a result their prescriptive pronouncements often conflict with social reality such that when the rubber of public policy hits the road, it quickly burns out because of excessive friction.
It would be wonderful if we could experiment with new ways of doing public policy that worked with, rather than against, our collective identity. To do that though, we would first need to acknowledge our collective ignorance about a great many things.
UPDATE: It has been reported this morning that self-reported hunger in early March this year has gone up to the highest level it has ever been (exceeding the previous peak in 2008).