The World Economic Outlook (WEO), brought
out by the IMF in April 2003, contains detailed chapters on several issues-on
'Why Bubbles Burst', 'Unemployment and Labor Market Institutions: Why
Reforms Pay Off', 'Growth and Institutions'. This review intends to focus
on the last of these-first, because it is a crucial and potent issue that
has severe repercussions for the future of developing economies, and second,
because it questions issues more interesting and perplexing than the IMF
would have us believe.
The overall conclusion of the chapter on 'Growth and Institutions' is
that institutions play a major role in determining per capita income,
economic growth and growth volatility across countries. And all the variables
used by the report to measure institutions support this contention. In
addition, the report finds that across different regions of the developing
countries, per capita income and institutional quality rise in tandem.
The pattern is not so consistent in the case of differences in growth
rates and growth volatility. However, the question of causality is much
more tricky, as is admitted by the report. The analysis is definitely
very detailed and is set within the context of the literature although
there are some interesting omissions.
The report introduces the concept of 'perceptions and assessments' of
public institutions as explanatory variables. It uses three measures of
institutions. To quote, 'these indicate, first, the quality of governance,
including the degree of corruption, political rights, public sector efficiency,
and regulatory burdens; second, the extent of legal protection of private
property and how well such laws are enforced; and third, the level of
institutional and other limits placed on political leaders' (WEO, p. 3).
The perceptions approach is used on the grounds that 'perceptions of the
political, economic, and policy climate embodied in the institutional
measures are likely to be of key importance in shaping overall conditions
for investment and growth. Given the mobility of international capital,
for example, such assessments may play a major role in determining a country's
ability to attract and retain investment inflows' (WEO, p. 3). The role
of governance is treated as a key institutional variable. Its definition,
boasts the report, is much wider and covers not just the extent of corruption
as is sometimes seen in the literature. Recent literature has however,
included many of these factors. The exact indicators have actually been
used from other studies.
The line of enquiry is not new, nor are the conclusions reached. For quite
some time now, the role of institutions in determining economic growth
and incomes has been a hot topic for debate. Finding its roots in the
writing of Douglas North, a historical account of the issue is to be found
in David (1994). Many empirical analyses are to be found in the literature
e.g. Easterly and Levine (1997), Barro (1991), Knack (1996), Knack and
Keefer (1995). The recent debate has concentrated on a few major views.
The first, forwarded among others by Jeffrey Sachs, John Gallup, Andrew
Mellinger and Jared Diamond, is that geographical and historical factors
like climate, soil fertility, location, climatic environment and natural
resource endowments play the determining role in deciding the level of
economic development. This view underplays the role of institutions in
such determination.
A second view, proposed among others in the works of Acemoglu, Johnson,
and Robinson (2001, 2002), Engerman and Sokoloff (1997, 2002) and Rodrik,
Subramanian, and Trebbi (2002), is that institutions subsume the effects
of historical and geographical factors and thereby play the decisive role
in determining the path of economic development and growth. For example,
for all ex-colonies, the institutional framework will be shaped, at least
in part, by the history of its colonization – its nature and extent.
That is why, this line of argument suggests, European established settler-colonies
like the US, New Zealand or Australia have well-established institutions
that support property rights, enforce the rule of law, and hence support
investment and growth. This line of opinion holds institutions as the
key determinant of economic well-being.
A major branch of empirical research in this area concentrates on the
effectiveness of the particular forms of institutions in determining the
path of development (see for example, Dani Rodrik, 2000). While some have
found financial institutions to be the driving factor, many others point
towards political institutions like the role and the stability of the
government, democratic elements within the political system and the effect
of corruption. The role of the state comes in for a major critical analysis
here. Accusations of 'crony capitalism' or corrupt governmental and other
machinery (like banking institutions) fostered by the state have been
regularly hurled by the Washington Consensus to explain the Asian collapse.
This opens up questions as to which forms of institutions should be taken
to be the pivotal ones, and whether the role of one form comes into conflict
with that of another.
While it is undeniable that institutions have a major role to play in
the development of a country, a crucial problem remains with specifics.
The first is the choice of institutional variables, from which the study
is obviously suffering. Secondly, choosing the exact quantitative or qualitative
measure poses another problem. The need to distinguish between descriptive
measures such as absence or presence of a particular type of institution,
and the quality or performance of that institution is very important.
The third problem is of data and methodology. This has also affected the
study but to this we would come later. The fourth is the technical problem
of using reduced form growth equations as opposed to structural growth
equations, which much of the cross sectional studies have actually done.
Fifth, institutional variables may be endogenous and not exogenous which
would cause serious problems for an econometric analysis. This has been
pointed out in the report but not taken care of. (for a reading of the
type of problems that may arise with such studies, see Aron, 2000)
Right at the beginning, the WEO's use of the concept of 'perceptions and
assessments' introduce some arbitrariness into the measures of institutional
quality used by the study and raises some doubts regarding the results.
Though it claims to have used aggregates of over 300 indicators including
ratings by country experts and survey results, the arbitrariness remains
as is admitted by the report in the next line itself (Footnote no: 32,
page 26, chapter 3, WEO, April, 2003). Surveys are very often biased and
unfortunately, so can be experts' opinions.
The fact is that no measure can quantify institutional qualities easily.
Measures of the quality of institutions that affect economic exchange
are suspect themselves. Clubbing various measures are also problematic.
The literature on economic growth typically has classified and treated
the proxies collectively as "sociopolitical measures." This
practice has tended to obscure the different channels through which institutions
operate and has impoverished the interpretation of the role of institutions
in growth. This is a serious flaw in analysing developing countries, especially
since where 'weak' institutions are implicated in low growth. The IMF
report suffers specially from this deficiency.
In addition, the IMF is completely preoccupied with variables with respect
to government inefficiency and corruption, as well as political freedom.
In fact the 'quality of governance' variable, includes as a major factor
'regulatory burden' which is a measure of the 'absence of government control
on goods market, banking system and international trade'. The measure
therefore places undue emphasis on the so-called freedom of the country
from government controls.
This emphasis on political powers of the government is deliberate. Positive
association with growth and income levels then enables the IMF to grossly
interfere with domestic political systems and workings of the government,
and force the state to withdraw in the political as well as the economic
sphere. And this is a necessity for the IMF, not for improving the lot
of the developing countries, but to establish the dominance of private
finance capital and multinational giants that it protects. This is the
same task that is being undertaken by US military control of 'rogue forces'
like the Taliban and Saddam Hussain in Afghanistan and Iraq.
At the same time, there seems to be an absence of most other types of
indicators considered in the literature. To name just a few of those mentioned
in the literature, indicators of business risk, social capital, local
government performance, ethnic and racial tensions, social development
and capability (including literacy, social mobility, size of urban classes
and crude fertility rate), are all absent from the analysis (for a detailed
discussion, see Aron (2000)).
In particular, 'economic institutions' in the form of banking and financial
institutions are glaringly from the analysis. While these are not taken
to be typical parts of the Anglo-Saxon system of capitalism, these are
very much a part of 'Continental capitalism' which is based on the prominent
role of banks in corporate finance and control. Many developing countries
in fact fit this model of capitalism better than the Anglo-Saxon one which
has obviously been the basis and the limitation of the WEO. In fact, there
is a pretty large literature that discusses the role of banking institutions
in determining the path of economic growth and development, especially
in transition economies (for a comprehensive discussion, See Cernat, 2002).
But perhaps this is more than an oversight or an inability to recognize
forms of capitalism other than the dominant one. It is more likely a denial
of the powers of domestic banking and financial institutions that have
been systematically weakened by the financial liberalization policies
pushed by the IMF itself. That would conveniently fall into place with
the rise to dominance of the US Central bank as the single powerful financial
institution in the world.
The 'role of governance' factor is very strongly related to growth, income
levels and growth volatility, finds the report. Judging by its definition,
the value of this variable should be high for a democratic country, but
cases of South Korea and China prove to be strong examples to the contrary.
In fact, China does not even possess well-defined property rights, which
is in fact the second institutional variable that the report considers.
Nor does it possess the third, namely, institutional freedom of its political
leaders. There are many other examples that defy this result. Even countries
like Saudi Arabia, which has had a history of a government ruling with
American military support, has a high income just because of its oil.
Here, the factor of geographical or historical endowments would turn out
to be more important.
A very important question arises when we come to the discussion on the
role of policies viz.a.viz institutions. The policy variables include
indicators involving inflation, trade openness, exchange rate over-valuation,
government size, financial development and capital account openness. The
report finds that the effect of policies (mainly involving trade openness,
stronger competition and higher transparency) on per capita incomes across
countries becomes insignificant when institutions are accounted for. The
latter turns out to be the determining and significant variable. Only
in two cases are the policy variables significant. One is where the financial
development variable is found to have a significant impact on growth.
The other is where exchange rate overvaluation is found to increase volatility.
Similar results have also been found by many other studies, notably by
Rodrik (2002).
Now, this turns out to be rather an embarrassment for the IMF-World Bank
since they have relentlessly advocated policies of macro economic management
and openness to trade as key forces for economic development and higher
incomes. World Bank economists like David Dollar and Art Kraay have churned
out paper after paper to prove that this is so. Works of J. Barro (1997),
X. Sala-i-Martin (2002), have also publicized on a mass scale, similar
findings. Of course, these did not take into account institutional factors
as such but much of the later IMF justification of its policies viz.a.viz
growth, poverty and income inequality have depended heavily on these results.
Caught in this predicament, the Report has desperately tried to save the
situation by pointing out, how this can be explained (page 10,16, Chapter
3, WEO, April 2003). Many explanations are offered. First, last forty
years of policies (captured in the measure used) cannot properly reflect
the impact on present incomes, which may be a result of a period of policies
much longer than that. Second, policies cannot be implemented properly
and their effects on growth are weak unless proper institutions are in
place (page 10, Chapter 3, WEO, April, 1993). Third, the directions of
causality between institutions and policies are not clearly known and
works, in most cases, in both directions (page 7, Chapter 3, WEO, April,
1993). Fourth, policies are part of the institutions themselves and may
be correlated heavily with the institutional variables which would render
them insignificant in a regression exercise. Fifth, Statistical and econometric
analysis based only on cross sectional data (used by the study) that catches
long term effects is unable to effectively capture the impact of policies
which can change very rapidly. The fourth and the fifth explanations in
particular, however show that there are major problems with this kind
of an analysis. In this case, we cannot take any co-efficient at its face
value.
The result viz.a.viz policies certainly raise another question. Policies
of increased financial liberalization, or capital account openness (as
the specific variable used by the report), have led to disastrous results
in many economies, notably those in many countries in Latin America, Asia,
and in Turkey and Russia. The main reason has been the volatility in short-term
capital flows. It has been admitted as much by IMF and World Bank officials
- for example, by the Economic Counselor and Director of Research of the
IMF, Michael Mussa (2000)(see also Global Development Finance, 2001, World
Bank, Chapter 3). Shouldn't the results found by the WEO, if tested correctly,
have shown a negative effect of financial openness on growth and, or at
least a positive one on growth volatility?
In the concluding section to the discussion on institutions and policies,
the Report ironically points to the weaknesses of its own methodology.
To quote, 'the "bottom line" from these findings is not that
policies are unimportant, but that our econometric framework (which is
constrained, in particular, by the limited time series data on institutions)
is not well suited to uncovering a relationship between policies and growth
that may well be revealed through time' (page 17, Chapter 3, WEO, April,
2003).
If we are to believe them, we cannot take any of the conclusions seriously.