The (mis)use of ICAs

If the term ICA does not mean anything to you, you have not been paying much attention to development policy in recent years. The acronym refers to Investment Climate Assessment, and it has been the latest rage at the World Bank. It is essentially a survey of enterprises that asks a host of detailed questions about both the characteristics of their operation and the main constraints they face. The Bank now has a new Investment Climate Unit that maintains and analyzes these surveys, which have been carried out in more than 30 countries.

In the words of the Bank:

Work to improve the investment climate is recognized as a key pillar of World Bank Group efforts to promote economic growth and poverty alleviation in developing countries. ICAs provide a standardized way of measuring and comparing investment climate conditions in a country; replacing a number of varying and sometimes ad hoc methodologies of the past. They are envisioned by the World Bank Group’s Private Sector Development Strategy as a systematic means to “allow i) better identification of the features of the investment climate that matter most for productivity and hence income growth, especially for poor men and women, ii) tracking of changes in the investment climate within a country, and iii) comparison of countries or Regions within countries.”

As the quote makes clear, these surveys are used increasingly to diagnose the main constraints facing firms and to identify policy reform priorities. If, for example, firms in country A complain most about the cost of finance while in B they complain about a skill shortage, this is taken as an indication country is constrained by poor access to finance while country B is constrained by poor human capital.

Sounds reasonable enough, right? Unfortunately, it’s not so straightforward. Inferences of this kind are quite likely to be wrong, and policy priorities based on what an existing sample of firms reports as major headaches are almost certainly biased.

To see why, you need to think in general-equilibrium terms and understand that the type of firms and operations that exist in an economy already reflects that economy’s constraints. In particular, the more binding the constraint with respect to a certain trait, the less likely it is that firms that depend on that trait will survive and prosper. And those firms that populate the sample will be those that either do not depend on that trait or have found some special way of getting around it. If I have little need for, say, public infrastructure, I am unlikely to report that its sorry state is a big deal for me. After all, if I depended on public infrastructure, I would not have lasted long. And this even if public infrastructure is the true binding constraint for the economy in the sense that a major improvement would unleash significant new entry and investment.

Or think of it this way. Whether the lack of X is a big headache for firms depends both on the severity of the constraint associated with X in general and on the importance of X to existing firms. As the constraint becomes more severe, the first factor becomes more important while the second becomes less. The net effect is ambiguous.

You may think this is all academic hairsplitting, but it’s not. Consider South Africa for example. A lot of people think that one of the most severe constraints that the economy faces is the high cost of labor (in dollar terms). This constraint finds reflection in a very high unemployment level and the weakness of the non-resource based tradables sector. Yet when the World Bank carried out an ICA recently, it found that firms complained mostly about lack of skills and about macroeconomic instability.

I think the explanation is that the structure of South Africa’s private sector already reflects the fact that one does not make money by operating a labor-intensive activity. So instead what you have is a bunch of skill- and capital-intensive enterprises in finance and other service industries, as well as in manufacturing, complaining about what matters not to the economy as a whole but to them specifically.

The bottom line? Surveys can be immensely useful if used intelligently, but identifying binding constraints requires a lot more than just asking firms.

Comments

[catkins]

Dani,

I enjoyed this post very much. It seems to me that survey results are likely to reflect what is concerning firms at the margin. This contrasts with your assertion that the binding constraint is that facing the economy in general and firms in total (as opposed to marginal; the lexicon here is not clear to me).

From here though I would turn the table back to you: Is it at the firm or economy-wide level the World Bank should operate? Is it at the marginal or total level?

I think this is a complex question as the solution invariably depends upon the political climate, nature of the constraints, and the institutional capacity to change. I would instinctively suggest that more success would be found at the margin and for this reason the ICA's might indeed by very useful.

[Dani Rodrik]

catkins --

there is nothing very complicated about this in principle. the reforms should seek to have the maximum impact AT THE MARGIN for the economy as a whole. The point is that you cannot tell how much the desert will benefit from water by asking camels.

[catkins]

Dani,

Thanks for your response. I understand your argument. I am interested in how the value of the information should be decided. I believe that this value is created by its purpose. Therefore, I am interested in the types of reforms the World Bank is seeking to perform.

For instance, in South Africa, should labour market reform (through reform of labour laws, seeking to reduce the influence of COSATU et al) be the priority or should skills development be put first. Should the binding constraint be addressed, a move that is likely costly, difficult and in which success may be less certain. Or should the marginal complaint be addressed, a area in which there may be 'quick wins' available?

Perhaps this is too far a digression from the post. Hopefully, my stocking will contain a copy of your recent book to develop my understanding of this.

Thanks,

[Abdi Hassan]

You get more useful information from people wishing to start enterprises (entrepreneurs)than the pre-existing firms which are bound to act in their self interest.

[Dominic]

Catkins,

You're confounding the issue. The binding constraint is the one which, upon removal presents the largest marginal gain at lowest marginal cost.

You seem to be seperating the idea of marginal constraint from binding constraint, but you should think of these as the same thing.

What you should think of as different, is that the binding constraint for a particular firm may be different to the binding constraint for an economy.

The firms management is responsible for solving its own binding constraints. Government is responsible for removing binding constraints of the economy.

In the example given, Dani's point is that the firms which exist in S.Africa think the binding constraint is skills, because those firms which have survived in S.Africa are not producing labour intensive products. But, according to this example, reducing the cost of labour would have greater effects in S.Africa than increasing the supply of skills. The difference is because the firms sampled have been selected as only those firms which have survived in a high labour cost environment. Reducing the cost of labour would lead to the entry of new economic activities, but that can't show up in a survey of existing firms.

[Chandan]

Nice post...true, very true...but what can be done to make sure that such surveys reflect true binding constranits of the economy? or are these kinds of surveys the best option among the worst?

[Robert]

Dani,

You bring up a good point, but I would have to say you are critizing a "second best" solution.

Do you work on what is binding a country's existing industries or do you work on those constraints that are keeping new industries from emerging in the first place?

As a person, do you focus on your strengths or do you focus on your weaknesses?

The answer is there is no right answer. It depends on the person. It depends on the country.

These surveys are good start and very valuable, but as you point out they are missing a whole set of constraints which may be essential. A truly optimal decision would require more than these surveys provide.

The problem is we can easily identify those constraints on our strengths and we are comfortable dealing with those more so than we are our weaknesses. A second best effort focuses on identifying what is easy. A first best effort looks at the whole picture..., but that is exponentially harder...

[inthemachine]

I wonder if there are some relevant lessons from research on availability bias.

[Per Kurowski]

Rodrik: “Surveys can be immensely useful if used intelligently, but identifying binding constraints requires a lot more than just asking firms.”

Rodrik is absolutely right! Especially when some of the most severe binding constraints are hard to confess… such as an ingrained cultural trait like the lack of trust and of the owners not wanting to give up any sort of control, keep the enterprises from developing their full potential.

[paine]

by passing
the general point
well made as usual
and off tracking on the example

are you suggesting
south africa oughta lower its forex rate
like them asians done

[paine]

"survey of enterprises "

are these "enterprises"
local outfits
or are they more often
TNC tie ins of some sort ??

if as i conjecture
IMFers are concerned
about what makes a place useful to the TNCers
then its strickly
cherry picking anyway

[Vijaya Ramachandran]

I see your point--it is certainly the case that existing firms endogenize some of the constraints. But if this is always the case, the complaints would not correlate with external, objective measures of the business environment (or the investment climate, as the Bank calls it). One can test this...and we have tried. My paper ("What Matters to African Firms? The Relevance of Perceptions Data," with Alan Gelb, Manju Shah and Ginger Turner) finds that perceptions are helpful and even camels might sometimes complain about the lack of water in the desert....

http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2007/12/14/000158349_20071214152644/Rendered/PDF/wps4446.pdf

We'd love to hear what you think! Regards, Vijaya

[paine]

the world bank
is
tiny tim
as played
by the imf scrooge

[tutti-frutti]

These and related issues were acknowledged, reviewed and assessed by the WB itself some years ago in its World Development Report on the investment climate (2004?). What's new?