Thursday, December 10, 2009

The Impact of Governance on the Economy - Part 1, The Macro Story


Many scholars and practitioners study the linkage between governance, corruption and economic development. In October 2009 the Lodewijk Meijer Group presented some preliminary findings of one of our latest research projects at a conference entitled 'The Economics of Corruption'. The conference was organized by the University of Passau in Germany. The paper was authored by your author, Erik L. van Dijk, and co-principal Margaret Rose. The theoretical story is straightforward. Corruption and bad governance hinder economic development. However, empirical results are not really convincing. Why is that so?

In our paper we constructed a Governance Ratio based on indicators from Transparency International, the World Bank, and the World Economic Forum. Governance deals with issues like corruption (including corruption perception), bribery, legality (presence and quality of a robust legal system), judicial efficiency, political factors, shareholder and other rights, et cetera.

Countries were classified into one of five groups. Since it was already clear from earlier research that there is a linkage between governance and economic development when looking at a country's level of GDP per capita (absolute analysis), we decided to look at developed and developing nations separately. The developed block was divided into a 'good' and 'poor' governance category (quintiles 1 and 2). The remaining, developing nations were categorized into three blocks. Quintiles 3, 4 and 5 were labeled 'good', 'average' and 'poor' governance. Graph 1 shows that the governance ratios of 'poor' developed nations are still - on average - higher than the 'good' scores of 'good' developing nations. Result: when moving from quintile 1 to quintile 5 we seem to be moving from 'high' to 'low' in a neat, linear fashion.

Graph 1: LMG Governance Ratios

We proceeded by analyzing the relationship between the governance quintiles on the one hand and economic growth on the other. Graphs 2a, 2b and 2c present the results. Graph 2a uses all available data in the LMG research database from 1980 onward, with economic growth rates based on IMF data. In graph 2b we look at the last 10 years of data and in 2c we focus on growth forecasts (up to 2014). Again, economic growth data are from the IMF. With averages in some cases being distorted by outliers in some (smaller) countries, we present median results and not arithmetic averages.

  Graph 2a; Governance and LT Economic Growth
IMF GDP real growth rates 1980-present and LMG Governance ratios

The long-term analysis shows that developing nations did indeed grow faster than developed ones during the period 1980 - present. However, the difference in real economic growth rates is rather small when taking into account that population growth rates in developing nations are also higher. And not just that, when looking at absolute growth in money terms, the gap between rich and poor has continued to grow with the only exception being some fantastic development stories in a few successful emerging countries. When looking within the developing group, results are quite interesting. The 'good' governance group outperformed the 'average' and 'poor' quintiles (compare q3 with q4 and q5). When taking into account that population growth was actually higher in q4 and q5 than in q3 this is an interesting finding. But nevertheless differential growth rates weren't that big. And when looking at the last 10 years - i.e. the period of ongoing globalization - we see a remarkable flattening of the graph when comparing graph 2b below with graph 2a.

Graph 2b; Governance and Econmic Growth - Last 10 years
IMF Real Econ Growth data and LMG Governance Ratios

The relationship between governance and economic growth within the developed world is once again not very clear. But this time we see that it is also not so clear anymore within the developing nations. The average growth rates in these nations went up, due to the combined effect of it being a relatively good period for them (notwithstanding the Global Crisis, but in the 1990s they had the Ruble Crisis, Asia Crisis and Mexican Peso Crisis to deal with!) and strong growth in foreign domestic investments (FDI) by Western investors on the one hand and increased portfolio investments in their equity and bond markets on the other. But the relationship with governance was at best unclear.

The Global Crisis was of course an interesting period that will undoubtedly have a major impact on economic development in the world. Graph 2c does therefore present the linkage between governance quintiles on the one hand and forecasted economic growth rates on the other.

Graph 2c; Governance and Forecasted Real Economic Growth
IMF Economic Growth Data and LMG Governance Ratios
Note: Expected growth rates for the years 2009-2014

Graph 2c shows clearly the impact of the Global Crisis and contagion effects between markets. The expected growth rates in the developed world and in q3 (developing nations with relatively good governance) are now much less high than those of the nations in q4 and q5. It remains to be seen if this is somehow also related to longer-term effects. For the time being we at LMG believe that the short-term explanations that link this phenomenon to the Crisis are more plausible. However, longer-term (more structural) factors that deal with demographics and the availability of scarce resources could play a role. Time will tell. 

Rose and van Dijk did also look at the linkage between Governance and other economic indicators in the LMG report. Since theoretical studies about governance and economic development did all indicate that good governance would attract more foreign capital, the authors concentrated on indicators of international economic linkage in particular. The authors looked at 1) Trade Openness( i.e. Imports and Exports); 2) Foreign Reserves and Gold; 3) External Debt; 4) Market Value of Equity; 5) Foreign Direct Investments (at home and abroad) and 6) Foreign Aid.

It would go too far to go in detail here, but the general conclusions were:
  • There is a positive linkage between Governance on the one hand and Trade Openness on the other. A risk factor for developing nations is that it seems to be easier to expand imports than increase exports. In other words: when good governance helps establish a better base for increased international contacts / exposure discipline is required so as to not create a foreign currency drain via trade deficits.
  • Developing nations have relatively larger Foreign Reserves and Gold, when using GDP as the scaling factor. Their financial system is less developed (less developed banking and insurance system which offers less products; lower leverage) with foreign currency reserves and gold automatically playing a larger role. The Global Crisis indicated that this is not just plain 'bad'. On quite some occasions leverage in the Western world went too far. That is of course not to say that financial system improvements in the developing world aren't of the utmost importance.
  • External debt: part of the 'leverage' story above is also that developed nations are better equipped to borrow money internationally. Their economies are more developed and stable, their currencies stronger et cetera. Developing countries with better governance do have better opportunities to borrow internationally. However, earlier research by other scholars has indicated that the positive effects of international financial flows are least related to foreign debt. Equity portfolio investments and Foreign direct investments are far more healthy for the receiving country.
  • There is a clear linkage between the relative size of the stock market in a country (compared to the size of its GDP) and governance. Better governance attracts more portfolio investments.
  •  There is a clear linkage between good governance and Foreign Direct Investments (FDI). A better governance track record helps attract more foreign direct investments. Earlier research indicated that FDI is the most important (in terms of growth impact) international flow indicator, with portfolio investments a good second. (Bank) Debt-related flows are least important since these flows can be reversed relatively easily, thereby increasing the risk profile of the country.
  • Last but not least: when looking at International Economic Aid there is a linkage with governance as well. One would expect the poorest nations to receive most aid. There is a linear relationship between wealth and our governance quintile distribution, with q1 being the richest category and q5 the poorest. However, q5 is not the group of countries that receives most aid when comparing the absolute amount of aid in US dollar terms to the size of the economy! The 'average' governance group, q4, scores best in this respect. 
Dead Aid - Dambisa Moyo

But it remains to be seen if it is that 'good' or 'important' to do well in governance terms for the purpose of receiving economic aid. Dambisa Moyo Ph.D - Zambian economist, formerly Goldman Sachs and author of the book 'Dead Aid' - has done research that indicates that 'more aid' is not necessarily good for the receiving country. It might also be in the interest of the donor, helping the latter to keep the former 'dependent'. Two observations are important: 
  • Economic Aid is always relatively small compared to the relative size of the other economic flow factors (FDI, portfolio investments). True, it is often received under 'better' conditions (free when it is a gift or at a lower interest cost in case of loans), but even when correcting for this difference it is still relatively less important. 
  • When looking at the relationship between Trade Openness and Governance, q4 is the quintile with the most shocking deficit. The deficit represents a net money flow from the 'poor' to the 'rich' when keeping in mind that the bulk of imports from q4 countries comes from richer countries with the bulk of their exports going this direction as well.
Graph 3 shows the size of received economic aid per quintile and the net trade surplus/deficit.

Graph 3; The Relationship between Governance, Trade Surplus/Deficit and Economic Aid
Data: Lodewijk Meijer Group, IMG and CIA Factbook (2009)

We do therefore agree with Dambisa Moyo that the potential dangers of receiving aid far outweigh the benefits, especially in the longer run. With the real dangers going much further than just a mere trade deficit. Aid can make economies 'lazy', trigger 'corruption' with leaders that received a 'seal of approval' by the donor country (and often not by their own population!) first extracting a considerable amount of aid money for themselves with the economically more dynamic middle class finding it hard to initiate robust economic growth. Feudalism and dependency remain.

Governance as a Risk Factor
All in all, the LMG Research does corroborate earlier findings that the empirical relationship between economic development and governance is far less clear than the theoretical one. We do find a relationship, but it is worrying to see that it became less clear in the most recent period. Globalization did not have a positive impact. Isn't that strange? Most scholars and practitioners looked at Governance from a 'return' point of view: the linkage between GDP (or GDP per capita) or GDP growth rates and governance, et cetera.  In graph 4 we present the relationship between our governance quintiles on the one hand and the standard deviation of real GDP growth rates on the other. The long-term historical averages, the last 10 years and the forecasts (compare graphs 2a, b and c) are now combined into one graph when looking at this risk-related indicator.

Graph 4; Governance and Macroeconomic Risk
Risk measured as standard deviation of GDP Growth Rates
Data: LMG and IMF

The performance of countries in quintiles 1 and 2 (developed nations) was far more homogeneous when looking at the standard deviation of their real GDP growth rates. This is true for all three time periods (long-term history, last 10 years and the 2009-2014 forecasts). With the exception of the 2009-2014 forecasted period we see a linear relationship between the standard deviation of GDP growth and governance ratio in developing nations. The period 2009-2014 is probably different due to the impact of the Global Crisis. Contagion between markets had the strongest effect in those markets that were to a larger extent part of the global financial community. But it is important to note that part of the 'governance' story is not a return story, but risk-related. The group of countries with a poorer governance track record might on average do almost as good as the ones with a better track record when looking at GDP growth, GDP per capita, absolute GDP et cetera when analyzing developing nations. But: the standard deviation of these 'return'-related indicators is higher. 

In other words: the poor governance countries as a group are riskier. When combining the return and risk side of the equation we can still confirm that good governance is of value to a country. But we do now also understand why it was so difficult to get this confirmed in earlier research. Focus has been on the 'return' side with 'risk' telling an important part of the story.

Who is to blame? 
But risk and return along the lines discussed above are not the whole story. Dambisa Moyo's analysis of Foreign Aid and its 'bad' impact on recipient countries brings us to another aspect that needs further research: the international relationship between developed and developing nations. In and of itself international integration is good. But quite a few of the graphs above do indicate that the linkage between economic development and governance did get blurred since the start of the globalization period.

One reason was the potential for 'mistakes' in terms of international trade with imports growing faster than exports. Developing countries should be aware of this and avoid stimulating 'Trading' firms that bring non-necessary luxury products from abroad into the country. Trade discipline is an important factor when initiating an international integration policy.

Another reason is the Aid trap. Leadership of the country might initiate some non-necessary, cosmetic governance measures so as to ensure more aid with that aid doing not too much for the country.

A third reason was not mentioned yet, but we want to address this issue now before concluding this Macro-focused article on the Impact of Governance on the Economy. It relates to a discussion that already made it into academic and professional journals in papers about the value of Transparency International's Corruption Perception Index (CPI).

Transparency International's CPI Index

The CPI Index has quite a track record and is well-known as one of the important indicators when analyzing global governance. One of its strengths is that it is available for almost all countries in the world. The 2009 CPI Index, which was published last month, contains about 180 countries. The index is based on a set of composite surveys measuring corruption perception in a country. Various scholars and practitioners were critical of the approach. Perception is a qualitative factor that is very sensitive to behavioral flaws. Even when trying to avoid this by combining different surveys amongst different groups of people, you still cannot totally rule out sensitivity to behavioral biases. Another factor mentioned by the skeptics was the fact that English-language countries seemed to have a more open information flow and were more integrated into the international definition of the body of knowledge that formed the basis of the CPI index. The lesser known the language, and the less people spoke English or another lingua franca inside that country, the larger the possibility of unreliable surveys since there was less of an objective international information exchange This may all be true, but there aren't many feasible alternatives. Quantitative, more objective indicators aren't developed yet and available alternatives do often suffer from other weaknesses like for instance a relatively small number of countries for which data are available.

Within the LMG Governance Ratio the CPI Index is one of the factors. When comparing the 2008 outcomes with the new 2009 results, we see a very high correlation: 0.99. In table 1 we present the top-5 and bottom-5 of the ranking.

Table 1; Top- and Bottom-5 of the CPI 2009
Source: Transparency International

The top- and bottom-5 aren't really surprising: developed nations in the top and developing nations plagued by internal and/or external political turmoil at the bottom. Of course we don't expect this ranking to be completely wrong, but it is important to look at the CPI index in more detail. The reason being that corruption - as an important aspect of overall governance - always has two sides to it. Especially when dealing with international corruption. The perception of corruption score of CPI is constructed in such a way that negative events are mainly recorded in the poor country where corruption happens. However, quite often the government, company or individual offering / paying the bribe is from a developed country. Logical, when you are far richer than your business partner or political colleague it is very tempting to get what you want by increasing the 'transaction cost level' a little bit through the payment of an extra. This extra is of course technically 'corruption', but why should we only blame the receiver for it?

Suppose that a big Western construction firm tries to win a government order with a total value of let's say USD 50 million in some African nation. When offering a minister or other decision taker from the African nation USD 1 million in return for getting the order, this is only 2 percent of the transaction size for the construction firm and even less when looking at total firm size and not just this order. However, it might easily be 20 times an annual salary for the minister. In case of a similar bribery situation in a Western nation a minister might have a salary of USD 250,000 which translates the same 'deal' into just 4 times annual salary. When adding to that the relatively more transparent and independent press in the latter country - which increases risk for the bribed person -  it is far less likely that a receiver in a Western country will 'fall' for the same type of bribe. But: isn't that therefore to quite some extent a negative event that the donor should be blamed for as well?

This is an important caveat that we should make when analyzing CPI data and/or any other governance-related indicator. It also explains why Globalization seems to have made things not better but worse. In many instances Globalization led to an increased contact between 'rich' and 'poor' countries, i.e. an absolute growth in the number of potential corruption-sensitive transactions! Good governance is a matter of laws in both developed and developing nations, with the former having an important role to play in international organizations dealing with this issue. They should understand that risks that seem to be relatively small nationally, might easily translate into potential corruption calamities when looking at cross-border business deals involving developing nations.

This being said, it is nevertheless interesting to compare average CPI scores across different groups of nations. Table 2 presents an overview of some of these data. The first panel shows clearly that average CPI scores in developed nations clearly outperform those in developing nations. But within the developing nations differences between sub-groups are relatively small. Actually, the group of BRIC nations - Brazil, Russia, India and China, the strongest emerging economies - doesn't really stand out. And neither do the Next-11, a group based on Goldman Sachs's 2005 research into the potentially strongest next layer of growth economies.

Table 2; CPI Index Averages 2009
Various Sub-Categories

When looking at the geographical panel (last column) we see that Europe is the only area with a score larger than 5.5 on a scale from 1 to 10. This does to some extent reflect that governance policy is further advanced in Europe. However, it also reflects that the ratio developed / undeveloped is higher in Europe. That the situation is worst in Africa is not surprising. This is partly related to internal African governance and leadership problems, but also a direct consequence of the lower wealth levels in Africa. The lower the wealth level, the easier it is for outsiders to 'score' with bribes. And not just that, the relative influence on African economies through the 'Dead Aid' scam - as reported by Dambisa Moyo - is higher as well.

Based on our research there is a linkage between macroeconomic growth and governance. However, it is far less clear than theory suggested. It is also partly a 'return' related story and partly risk-related. Bad governance doesn't immediately translate into calamities when looking at average economic growth rates. True, it might make a country more sensitive to some risk factors, while on the other hand helping it to avoid others. Globalization during the last 10 years has clearly shown that international integration is not always an easy-win story for developing nations. First, the international exposure can easily lead to disturbances of the balance of payments due to the fact that import growth can outweigh export growth. Second, gains from international capital flows are only positive in the longer run when they are related to FDI (foreign direct investments) or equity portfolio investments. International flows related to bank debt do have a tendency to revert at exactly the wrong moment for most developing countries. And last but not least: it can create dependencies that are only beneficial to the developed nation and not necessarily to the developing nation.

Bottom-line: governance indicators are important at the macroeconomic level, but a governance-based investment strategy cannot be constructed. The factor is simply not important enough as a stand-alone indicator. This is not to say that we should not continue to work on governance improvement since it is important in the longer run. But we should understand that part of that work should be international in scope with developed nations accepting that they do have a responsibility for alleged 'corruption' and 'poor governance' in developing nations.

When moving from the macroeconomic level to the firm level the big question is: does this macro-level analysis translate 1-on-1 into the micro story? In other words: what about differences in governance between firms within countries? We will address this issue in part 2 of this contribution (to be published later this week).

No comments:

Post a Comment