ALBERTO ALESINA

ALBERTO ALESINA

Alberto Alesina was born in 1957 in Broni, Italy and obtained his Laurea
from the Università Bocconi, Milan in 1981 and his PhD from Harvard
University in 1986. His main posts have included: Assistant Professor of
Economics at Carnegie-Mellon University (1986–8); Assistant Professor of
Economics and Government (1988–90) and Associate Professor of Political
Economy (1990–93) at Harvard University. Since 1993 he has been Professor
of Economics and Government at Harvard University.
Professor Alesina is best known for his contributions, in terms of both
theoretical analysis and empirical investigation, to the various forms of interaction
between politics and macroeconomics; and his influential work on
politico-economic cycles, the origin and implications of fiscal deficits, and
the relationship between political stability and economic growth. Among his
best-known books are: Partisan Politics, Divided Government and the Economy
(Cambridge University Press, 1995), co-authored with H. Rosenthal; Political
Cycles and the Macroeconomy (MIT Press, 1997), co-authored with N.
Roubini; and The Size of Nations (MIT Press, 2002), co-authored with E.
Spolare. His most widely read articles include: ‘Macroeconomic Policy in a
Two-Party System as a Repeated Game’, Quarterly Journal of Economics
(1987); ‘Political Cycles in OECD Economies’, Review of Economic Studies
(1992), co-authored with N. Roubini; ‘Central Bank Independence and Macroeconomic
Performance: Some Comparative Evidence’, Journal of Money,
Credit, and Banking (1993), co-authored with L. Summers; ‘Distributive
Politics and Economic Growth’, Quarterly Journal of Economics (1994), coauthored
with D. Rodrik; and ‘The Political Economy of the Budget Surplus
in the US’ , Journal of Economic Perspectives (2000).
We corresponded with Professor Alesina in March/April 1997.
Background Information
How did you become interested in economics and where did you study as an
undergraduate and postgraduate student?
In high school I was very interested in socio-political problems. I thought that
economics was the most rigorous of the social sciences. As an undergraduate
I studied at Università Bocconi in Milan, Italy [1976–81] and took both my
Masters degree and PhD at Harvard University [1982–6].
Which papers and/or books have influenced your research interests?
I have been much more influenced by ‘facts’ rather than by specific papers or
books and have always been very interested in the policy-making process.
The basic fact which has always impressed me is how different actual policy
making is from the predictions of models which assume social planners and a
representative consumer. I always noted how in politics discourse is about
redistributive conflicts while in most macroeconomic models distributive
issues are absent.
Is there any aspect of the Italian economy and/or political system that stimulated
your interest in the link between economics and politics?
Yes: the inability of government after government in Italy to address serious
fiscal problems has been an important influence. More generally, thinking
about Italy made me wonder about the relationship between political fragmentation
and economic performance.
Do you regard yourself as belonging to any identifiable school of thought in
macroeconomics?
No.
Keynes and Keynesianism
Was Keynes naive in assuming that economic policy should be, and would be,
carried out in the public interest?
If we can characterize Keynes’s view as such, I would say yes, although I
think that your question oversimplifies his view.
Alberto Alesina 569
In Roy Harrod’s 1951 biography of Keynes he argued that Keynes tended to
think of important decisions being taken by ‘intelligent people’ and gave little
consideration to the political constraints placed on this vision by ‘interfering
democracy’.
If one reads beyond the General Theory, for instance the pamphlet How To
Pay For The War, I think that one can see that Keynes was aware of the
subtleties of policy making. I agree that in his main scientific work he did not
consider the effects of political distortions on policy making. You can’t expect
everything from the same economist. He did quite a lot as it is!
Do you think that Keynesian economics, with its emphasis on discretionary
fiscal policy, has fundamentally weakened the fiscal constitutions of Western
industrial democracies?
No. In my opinion this point has been overemphasized. Italy, for instance, has
accumulated a very large debt for several reasons. The adoption of Keynesian
policies is not one of them.
Have Italy’s fiscal problems stemmed from its political system?
Yes, from its fragmented political system, over-powerful unions, lack of a
strong party committed to fiscal discipline and an overextended and entrenched
bureaucracy.
Why are some countries more prone to budget deficits than others? Why is
deficit reduction such an intractable problem?
In my paper with Roberto Perotti [1995a] we conclude that it is difficult to
explain these large cross-country differences using economic arguments alone.
Politico-institutional factors are crucial to understanding budget deficits in
particular, and fiscal policy in general. While the economies of the OECD
countries are relatively similar, their institutions, such as electoral laws, party
structure, budget laws, central banks, degree of centralization, political stability
and social polarization, are quite different. In a companion paper examining
fiscal adjustments in OECD countries [Alesina and Perotti, 1995b] we find
that coalition governments are almost always unsuccessful in their adjustment
attempts, being unable to maintain a tough fiscal stance because of
conflicts among coalition members. We also find that a successful fiscal
adjustment is best started during a period of relatively high growth, does not
raise taxes, but rather cuts transfer programme and government wages and
employment. Politicians and their advisers must stop thinking of just about
everything on the expenditure side of the government budget as untouchable.
What do you regard as being the most important contribution made by Keynes
to our understanding of macroeconomic phenomena?
570 Modern macroeconomics
The idea that aggregate demand policy matters because of lack of complete
price flexibility.
The Political Business Cycle
Prior to Michal Kalecki [1943] were there other economists who anticipated
the possibility of a political business cycle?
Not that I know of.
To what extent could the deep UK recession of 1979–82 be viewed as being
politically induced? For example, it can be argued that the then Prime Minister,
Margaret Thatcher, used high unemployment in order to restore the
health of British capitalism, a development consistent with Kalecki’s model.
Thatcher’s recession was the result of the need for disinflation. I am not sure that
unemployment is needed ‘in order to restore the health of capitalism’. Unemployment
is in part a cyclical phenomenon, in part the result of supply-side
rigidities. This latter influence is particularly strong in the European economies.
How important are macroeconomic variables for voting behaviour?
In the USA, the rate of GNP growth is extremely important, inflation and
unemployment somewhat less so. In other countries the evidence is less clear
cut, in part due to differences in the electoral systems. In general, the state of
the economy is very important for elections but how this effect manifests
itself may vary dramatically from one country to another.
Nordhaus’s [1975] model is intuitively very appealing. What do you regard to
be its main strengths and weaknesses?
Its main strength is that it makes a simple powerful point which is easily
testable. Its main weaknesses are that it is based on the assumption of extremely
naive behaviour, and also has very weak empirical support.
How can voters (principals) ensure that their agents (the politicians) refrain
from opportunistic behaviour which creates economic inefficiencies?
Mainly with the threat of voting them out of office.
Non-Rational Partisan Theory
What do you regard to be the important contributions of Hibbs [1977] to the
development of politico-economic models?
Hibbs played an important role by introducing ideological–partisan differences
and moved attention away from models in which it was assumed that
all politicians have identical motivations.
Alberto Alesina 571
How strong is the evidence that inflation is more harmful to higher income
than lower income groups?
The evidence on this is not very strong. How harmful inflation is depends on
the level of inflation and various institutional arrangements, such as indexation.
For the post-war period as a whole, is the empirical evidence supportive of
the partisan view that left-of-centre governments favour and achieve lower
unemployment than right-of-centre governments? Have left-of-centre administrations
also been associated with higher inflation than right of centre
administrations?
To some extent yes. However, the success of left-wing governments in reducing
unemployment has only been temporary. After taking into account
structural breaks such as the break in exchange rate regime in the early
1970s, the oil shocks and so on, it is true that left-of-centre governments have
been associated with higher inflation than those right of centre.
Rational Expectations and Business Cycle Models
Do you attribute the decline of interest in politico-economic explanations of
the business cycle between the late 1970s and mid-1980s to theoretical developments
associated with Robert Lucas and the rational expectations revolution
or was empirical failure a more important factor?
I think that the ‘rational expectations revolution’ was a much more important
contributing factor to the decline of interest in such models. Furthermore,
empirical ‘failures’ have not stopped economists investigating this matter
further in the 1990s.
An important criticism of democratic markets is that voters are, in most
cases, uninformed. For each voter the benefits of gaining more information
will be outweighed by the costs. In this world of imperfect information is it
not inevitable that politicians will engage in opportunistic fiscal behaviour
prior to elections?
To some extent this is the case, but I would not overemphasize this point for
several reasons. First, if this were the main explanation for budget deficits, it
should apply (more or less) to every democracy. Thus one should not observe
such large differences in fiscal policies in OECD democracies. Second, large
deviations from efficient policies, such as huge deficits in election years, are
easily observable, if not by the individual voter, then certainly by the press.
Third, I do not know of any conclusive evidence which shows that larger
deficits favour the re-election of an incumbent. I think that in practice what
happens all the time is that in election years fiscal favours may not be very
572 Modern macroeconomics
large, and may therefore be hard to detect. Nevertheless, the political benefits
may be quite significant. Let me add on this point that I am not a great fan of
models of politics where imperfect information is the critical factor. I personally
think that conflict of interest is much more important than asymmetric
information.
In 1988 George Bush told the American electorate: ‘Read my lips, no new
taxes’. Each year in the UK the Chancellor of the Exchequer in his budget
speech emphasizes the competence of the administration in managing fiscal
affairs. Do you regard these examples as typical of the kind of political
behaviour predicted by the Rogoff–Sibert [1988] class of rational opportunistic
models?
With the caveats discussed in the previous answer, I think that Rogoff and
Sibert have a good point. Their model is also much more consistent with the
empirical evidence than the original Nordhaus model. My empirical research
on the subject is quite supportive of their model. This research is included in
Alesina and Roubini, with Cohen (1997): Political Cycles and the Macroeconomy,
MIT Press.
What role does ‘fiscal illusion’ play in politico-economic models?
I think that this concept is oversold. Explanations of excessive fiscal deficits
based on fiscal illusion are not totally convincing because they imply a
systematic bias in the errors made by the electorate concerning their estimation
of the costs and benefits of taxes and spending. Fiscal illusion is also
unable to explain the timing of the deficit problem in OECD economies or
the cross-country differences in budget deficits. I prefer models based on
rational behaviour and expectations.
Rational Partisan Theories
How strong is the evidence against the median voter theorem?
It depends. For large elections I would not use it: in fact, its key implication is
that when in office all the parties do the same thing. This is clearly inconsistent
with the evidence even of predominantly two-party systems that exist in
countries such as the UK and the USA. In multi-party systems the median
voter theorem proves to be even less applicable. Therefore, for macroeconomic
and macropolitical research I would not use it. On the other hand, if I
want to study voting behaviour on one issue in a committee of five people,
then the median voter theorem is a good start!
In the UK Tony Blair’s ‘New Labour Party’ has been moving closer to the
position adopted by the Conservative Party on many economic issues. For
Alberto Alesina 573
example, on 20 January 1997 the Shadow Chancellor of the Exchequer,
Gordon Brown, promised the electorate that ‘New Labour’ would not increase
tax rates if elected. On 21 January 1997 Tony Blair promised not to
reverse the Conservative Party’s industrial relations legislation of the 1980s.
Does this imply that polarization is less important than it used to be?
In the case of the UK we shall have to wait and see! However, on the more
general issue of increasing party convergence, in Political Cycles and the
Macroeconomy [1997] we conclude that ‘the idea that political parties are
becoming more alike when it comes to macroeconomic management is somewhat
exaggerated’. Our view is that while the macroeconomic problems of the
1970s and early 1980s ‘have probably made politicians on both sides of the
political spectrum more cautious in terms of macroeconomic management,
they have not completely eliminated ideological differences’. Furthermore, we
also point out that ‘both left-wing and right-wing governments in the next
decade will have to face issues of fiscal retrenchment’ and that ‘partisan conflicts
are very likely to explode on how to achieve this goal’.
In your rational partisan models [1987, 1989] you appear to adopt an
eclectic approach to your theoretical framework based on a variant of the
‘monetary surprise’ rational expectations model associated with Fischer
[1977]. Lucas and other new classical theorists now appear to attach more
importance to real shocks as an explanation of post-war macroeconomic
fluctuations. Does this not undermine the basis of models based on monetary
shocks?
I do not like the view that macroeconomic fluctuations are due to either
supply or demand shocks. First, an economy can be subjected to both types
of shock. Second, real business cycle models have certainly not been an
empirical success. The rational partisan theory does not imply that real shocks
do not exist. In fact in empirical testing one may want to control for supply
shocks.
What are the essential predictions of the rational partisan theory and have
they been supported by empirical work?
The basic idea of the model is that, given the sluggishness in wage adjustments,
changes in the inflation rate associated with changes in government
create temporary deviations of real economic activity from its natural level.
At the beginning of a right-wing government output growth is below its
natural level and unemployment is above its natural level. The opposite is
predicted for left-wing governments. After expectations, prices and wages
adjust, output and employment return to their natural level and the level of
economic activity should be independent of the party in office. However, the
rate of inflation should remain higher throughout the term of a left-wing
574 Modern macroeconomics
government. These implications of rational partisan theory are consistent
with the empirical evidence particularly for a subset of countries with bipartisan
systems or with clearly identifiable movements from left to right and vice
versa. The rational partisan theory is less applicable, and in fact tends to fail
in countries with large coalition governments with frequent government collapses.
In the UK the Conservative Party gained power in the four elections between
1979 and 1992. President Clinton has also been re-elected. What does the
rational partisan theory predict will be the effect of the repeated re-election
of a particular party?
Strictly speaking, it depends on how unexpected is the re-election. If the
Conservative Party had been sure to win in the previous four elections, not
much should have happened to the economy. Inflation should have continued
to remain low and growth (ceteris paribus) should have been stable. President
Clinton was probably less sure of being re-elected. In this case the
model predicts that growth should slightly increase in the USA.
What is likely to happen according to your model if a left-of-centre party
gains office at the peak of an economic expansion?
According to the model, the left-of-centre party will do as much as possible
to avoid a recession, including increasing inflation. If we speak of the UK in
1997, one has to consider other issues which may influence the outcome,
such as European Monetary Union. This kind of issue is clearly not considered
in the model.
What are the main policy implications of your model for the conduct of
macroeconomic policy and the design of political institutions?
The model in its stripped-down version is ‘positive’ rather than ‘normative’,
thus it does not have any direct policy implications. However, the
model can be used in a normative direction. Let me give a couple of
examples to illustrate the point. First, the model suggests that independent
central banks, by insulating monetary policy from partisan influences, can
reduce the extent of both monetary and real variability. This point is formally
derived in my paper with Roberta Gatti [1995]. Second, as for political
institutions, the model points towards a trade-off. Proportional electoral
systems which result in coalition governments lead to ‘compromise’ and
policy moderation. This reduces partisan fluctuations and polarization, but
may induce deadlocks in policy making, particularly with respect to fiscal
issues. Majoritarian systems leading to two-party systems have the opposite
feature, namely more policy polarization but no policy deadlocks. Extreme
versions of the two systems are unlikely to be optimal although I regard the
Alberto Alesina 575
risk of excessive ‘governing by coalition’ to be a particularly significant
problem in the European context.
Which is likely to create the most instability: exogenous or endogenous
timing of elections?
Endogenous elections undoubtedly generate more frequent elections. Whether
that creates more or less instability is less obvious. Italy is a perfect example.
One may argue that in the post-war period Italy has been very unstable
because it has had numerous government changes and many early elections.
On the other hand, one may argue that at least until the early 1990s nothing
much ever changed, because the same parties and individuals were always in
office.
What are the main weaknesses of the rational partisan theory?
I guess I am not the right person to ask that particular question, but elsewhere
[1995] I have pointed out that the Achilles’ heel of the rational partisan
theory is that the mechanism of wage formation is postulated exogenously
rather than derived from optimal individual behaviour.
Would you agree that rational voters will make their voting decision based on
information relating to both the past performance of a political party as well
as expected future performance?
Past information has to be used to form expectations about the future. Even
voters who are forward-looking have to look backward to form expectations.
The question ‘are voters backward or forward-looking?’ is very misleading.
Furthermore, I find the research on this point often both confused and confusing.
A different, more useful, question is whether the voters use ‘efficiently’
the past information that they have to form expectations about the future. In
other words, the question is whether voters use ‘rationally’ their past information.
This is a difficult but well-posed question, which is similar to questions
raised in finance.
To what extent do opportunistic behaviour and partisan behaviour depend on
the confidence of an incumbent administration that it will be re-elected? Is it
the case that the more confident an administration is of re-election, the more
likely it is to behave ideologically?
Possibly, perhaps likely. I have not seen a ‘rational’ model making this point.
Bruno Frey and Freidrich Schneider back in 1978 made this point in their
‘non-rational’ models.
576 Modern macroeconomics
General Macroeconomic Issues
How influential has the literature relating to time inconsistency, credibility
and reputation been to the development of politico-economic models?
In my case these contributions have been very influential. I think that this
literature has been an important driving force behind the new wave of politico-
economic models in the 1990s, although it is fair to say that now this
literature has a life of its own.
Why do governments create inflation?
One important reason is to try to reduce unemployment. A second reason
relates to the financing of budget deficits. Also, when inflation becomes
entrenched in the system it is costly to reduce it. These reasons may apply
differently to different countries, and at different points in time.
What is the relationship between political and economic instability? How has
the UK managed to combine political stability with a relatively poor economic
performance in the post-war period? How does this compare to the
Italian experience?
The UK has been ‘stable’ but very polarized. Conservative and Labour governments
have had very different programmes. On the other hand Italy has
had the opposite problem, namely frequent government changes but with
always the same people ending up in office. The result for Italy has been a
lack of fiscal discipline. However, concepts such as political stability and
polarization are difficult to measure in practice.
Should central banks have goal independence as well as instrument independence?
What is your view of the contracting approach suggested by Carl
Walsh?
Legislatures should set, once and for all, price stability (defined as low
inflation, between 0 and 3 per cent) as the sole goal of monetary policy.
Central banks should do whatever they want to achieve that goal. I do not
know whether you want to call this instrument or goal independence. Carl
Walsh [1995a] has produced a good paper but I would like to see further
work in a more political direction. For example, how does Walsh’s contracting
approach deal with the fact that in the real world we have partisan and
opportunistic politicians, rather than benevolent social planners?
Are you in favour of monetary and fiscal rules?
My answer to the previous question covers my views on monetary rules.
Personally I am against balanced budget rules for national governments but
in favour of such rules for sub-national governments.
Alberto Alesina 577
Does reducing inflation from moderate rates, say from 10 per cent to 5 per
cent, bring any significant real economic benefits in the form of improved
performance with respect to employment and economic growth?
If inflation remained stable forever at say 10 per cent, with perfect certainty
and everything adjusted, it wouldn’t matter very much. But higher inflation
results in more variable and less predictable inflation, which is costly. Thus,
on balance I believe that there are benefits in reducing inflation.
Where do you stand on the issue of European Monetary Union? For countries
like Italy and the UK do the costs of Monetary Union not outweigh the
benefits?
The benefits of Monetary Union have been oversold. There are clearly both
pros and cons to such an arrangement. Italy has benefited from the Maastricht
target, otherwise it would have done even less to put its ‘house in order’.
However, this is not sufficient reason to join a monetary union. Although a
full answer to this question would require a whole article, my feeling is that
the economic arguments in favour of European Monetary Union are quite
weak.
In your view what are the most important lessons and policy implications to
arise from recent research into the interaction of politics and the macroeconomy?
First procedures, namely how policies are determined. This matters for the
outcome. Second, when thinking about ‘optimal institution building’ one
should not ignore conflicts of interests. Third, models based on ‘social planners’
cannot completely explain the empirical evidence and may be misleading
or useless if used for policy prescriptives.
What kind of political system is most conducive to macroeconomic stability?
For an OECD economy I would choose an electoral system with a majoritarian
emphasis. The American system of presidential–legislative checks and balances
has also worked quite well. Different electoral systems imply different
choices in the trade-off between moderation and gridlock. An English system
is probably at the extreme of the ‘no moderation but no gridlock’ scale. The
current Italian system is at the opposite end. Perhaps the US system is a
happy medium.
At the moment what research are you currently engaged in and what in your
opinion are the important areas of research which macroeconomists should
concentrate on in the future?
I am working on three main areas. The first area is the issue of the political
economy of major fiscal adjustments where I have already published several
578 Modern macroeconomics
papers with Roberto Perotti. I am also interested in what determines the
number and shapes of countries, that is, economic theories of secessions and
mergers and their relationship with factors such as geography and trade and
so on. For example, see my paper with Enrico Spolare [1997]. Third, I am
also researching into the effect of socio-ethnic fragmentation on the choice of
fiscal policy in US cities and localities. Personally, I find issues relating to
fiscal policy more intriguing than those associated with monetary policy,
especially as we know less about the former than the latter. I think that the
political economy of social security reforms and, more generally, reforms of
the welfare state will be the number one item in the policy agenda of OECD
countries during the next decade. The research of economists needs to keep
up with important events.