The Decline and Renaissance of Opportunistic and Partisan
Models
The 1970s were a turbulent time for the capitalist democracies as the ‘Golden
Age’ of low inflation, low unemployment and above average growth came to
an end. The stagflation crisis of the 1970s also brought to an end the Keynesian
consensus which had dominated macroeconomic theory and policy making in
the quarter-century following the Second World War. In the wake of the
monetarist counter-revolution, Lucas inspired a rational expectations revolution
in macroeconomics. By the mid-1970s models which continued to use
the adaptive expectations hypothesis were coming in for heavy criticism from
new classical theorists, as the hypothesis implies that economic agents can
make systematic errors. In market-clearing models with rational expectations
the assumption that economic agents are forward-looking makes it more
difficult for the policy maker to manipulate real economic activity. There is
no exploitable short-run Phillips curve which policy makers can use. Preelection
monetary expansions, for example, will fail to surprise rational agents
because such a manoeuvre will be expected.
The rational expectations hypothesis also implies that voters will be forward-
looking and will not be systematically fooled in equilibrium. According
to Alesina (1988), ‘the theoretical literature on political business cycles made
essentially no progress’ after the contributions of Nordhaus (1975), Lindbeck
(1976) and MacRae (1977) because of the ‘devastating’ effect of the rational
expectations critique. The Nordhaus model involves an exploitable short-run
Phillips curve trade-off combined with myopic voters. Once the rational
expectations hypothesis is introduced, however, voters can be expected to
recognize the incentives politicians have to manipulate the economy for
electoral profit. Given that US presidential elections are held on a regular
four-year basis, it is difficult to believe that rational voters and economic
agents would allow themselves to be systematically duped by the macroeconomic
manipulations of self-interested politicians. Moreover, it is difficult to
reconcile the predictions of the Nordhaus model with situations where monetary
policy is conducted by an independent central bank, unless government
can in some way pressurize the central bank to accommodate the incumbent
government’s preferred monetary policy (on this see Havrilesky, 1993; Woolley,
1994). However, Blinder, in discussing this issue, denies that political pressure
was an issue during his period as Vice Chairman at the US Federal
Reserve, 1994–6. In his experience the political influence on monetary policy
was ‘trivial, next to zero’, although he agrees that this was not the case during
the Richard Nixon–Arthur Burns era (see Snowdon, 2001a).
The Hibbs model also has major theoretical shortcomings, especially with
respect to the stability of the Phillips curve trade-off implicit in his analysis.
Remarkably, Hibbs makes no mention of the expectations-augmented Phillips
curve in his 1977 paper even though the Friedman–Phelps theory was a
decade old and by then a well-established idea, even among Keynesians (see
Gordon, 1975, 1976; Blinder, 1988b, 1992a; Laidler, 1992a). The assumption
of rationality also has implications for the Hibbs model. Since the output and
employment effects of expansionary and contractionary demand management
policies are only transitory in new classical models, the identification of
partisan influences on macroeconomic outcomes will be harder to detect (see
Alesina, 1989). Alt (1985) concluded that partisan effects are not permanent
but occur temporarily after a change of government.
In addition to the theoretical shortcomings of the early political business
cycle literature, the Nordhaus model also failed to attract strong empirical
support, with the econometric literature yielding inconclusive results (see
Mullineux et al., 1993). While McCallum (1978) rejected the implications of
the Nordhaus model for US data, Paldam (1979) could only find weak evidence
of a political business cycle in OECD countries. Later studies, such as those
conducted by Hibbs (1987), Alesina (1988, 1989), Alesina and Roubini (1992),
Alesina and Roubini with Cohen (1997) and Drazen (2000a, 2000b), also find
little evidence of a political business cycle in data on unemployment and GNP
growth for the US and other OECD economies. This also applies where the
timing of elections is endogenous (see Alesina et al., 1993). More favourable
results for the Nordhaus model are reported by Soh (1986), Nordhaus (1989),
Haynes and Stone (1990) and Tufte (1978), who found some evidence of preelection
manipulation of fiscal and monetary policy for the USA. In addition,
Drazen (2000a) distinguishes between empirical predictions that focus on policy
outcomes (inflation, unemployment, growth) and those that focus on policy
instruments (taxes, government expenditure interest rates) and concludes that
‘the evidence for opportunistic manipulation of macroeconomic policies is
stronger than for macroeconomic outcomes’. With respect to policy outcomes
there is more support for the opportunistic political business cycle theory
coming from post-electoral inflation behaviour than can be found in the preelectoral
movement of real GDP and unemployment. Drazen also concludes
that the evidence in favour of opportunistic manipulation of policy instruments
is much stronger for fiscal policy than it is for monetary policy.
A significant problem for partisan theories in general is the argument and
evidence presented by Easterly and Fischer (2001) that low inflation helps the
poor more than the rich. Inflation acts as a financial tax that hits the poor
disproportionately because they tend to hold more of their wealth in cash
relative to their income than the rich, whereas the rich are more likely to have
access to various financial instruments that allow them to hedge against
inflation. In addition the poor obviously depend more on minimum wages
and state-determined income payments that are not always indexed to protect
against the effects of inflation (see Snowdon, 2004b). Easterly and Fischer
present evidence drawn from an international poll of over 30 000 respondents
from 38 countries and the responses indicate that the poor themselves are
more strongly averse to inflation than those with higher incomes. This undermines
a key assumption of Hibbs’s model that the rich are more inflation-averse
than the poor.
By the mid-1980s the politico-economic literature had reached a new ebb.
Zarnowitz (1985), in his survey of business cycle research, devotes one
footnote to the idea of an electoral cycle and refers critically to the ‘strong’
and ‘questionable’ assumptions of such models as well as the lack of supporting
evidence. The same neglect is also a feature of Gordon’s (1986) edited
survey of The American Business Cycle.
Following a period of relative neglect, the literature on the relationship
between politics and the macroeconomy underwent a significant revival since
the mid-1980s (see Willet, 1988). Economists responded to the rational expectations
critique by producing a new generation of rational politico-economic
models. Like the first phase in the 1970s, the second phase of politicoeconomic
models consisted of opportunistic and partisan versions of the
interaction between politics and the macroeconomy. In the next two sections
we examine the main features of rational political business cycle models and
the rational partisan theory.
Models
The 1970s were a turbulent time for the capitalist democracies as the ‘Golden
Age’ of low inflation, low unemployment and above average growth came to
an end. The stagflation crisis of the 1970s also brought to an end the Keynesian
consensus which had dominated macroeconomic theory and policy making in
the quarter-century following the Second World War. In the wake of the
monetarist counter-revolution, Lucas inspired a rational expectations revolution
in macroeconomics. By the mid-1970s models which continued to use
the adaptive expectations hypothesis were coming in for heavy criticism from
new classical theorists, as the hypothesis implies that economic agents can
make systematic errors. In market-clearing models with rational expectations
the assumption that economic agents are forward-looking makes it more
difficult for the policy maker to manipulate real economic activity. There is
no exploitable short-run Phillips curve which policy makers can use. Preelection
monetary expansions, for example, will fail to surprise rational agents
because such a manoeuvre will be expected.
The rational expectations hypothesis also implies that voters will be forward-
looking and will not be systematically fooled in equilibrium. According
to Alesina (1988), ‘the theoretical literature on political business cycles made
essentially no progress’ after the contributions of Nordhaus (1975), Lindbeck
(1976) and MacRae (1977) because of the ‘devastating’ effect of the rational
expectations critique. The Nordhaus model involves an exploitable short-run
Phillips curve trade-off combined with myopic voters. Once the rational
expectations hypothesis is introduced, however, voters can be expected to
recognize the incentives politicians have to manipulate the economy for
electoral profit. Given that US presidential elections are held on a regular
four-year basis, it is difficult to believe that rational voters and economic
agents would allow themselves to be systematically duped by the macroeconomic
manipulations of self-interested politicians. Moreover, it is difficult to
reconcile the predictions of the Nordhaus model with situations where monetary
policy is conducted by an independent central bank, unless government
can in some way pressurize the central bank to accommodate the incumbent
government’s preferred monetary policy (on this see Havrilesky, 1993; Woolley,
1994). However, Blinder, in discussing this issue, denies that political pressure
was an issue during his period as Vice Chairman at the US Federal
Reserve, 1994–6. In his experience the political influence on monetary policy
was ‘trivial, next to zero’, although he agrees that this was not the case during
the Richard Nixon–Arthur Burns era (see Snowdon, 2001a).
The Hibbs model also has major theoretical shortcomings, especially with
respect to the stability of the Phillips curve trade-off implicit in his analysis.
Remarkably, Hibbs makes no mention of the expectations-augmented Phillips
curve in his 1977 paper even though the Friedman–Phelps theory was a
decade old and by then a well-established idea, even among Keynesians (see
Gordon, 1975, 1976; Blinder, 1988b, 1992a; Laidler, 1992a). The assumption
of rationality also has implications for the Hibbs model. Since the output and
employment effects of expansionary and contractionary demand management
policies are only transitory in new classical models, the identification of
partisan influences on macroeconomic outcomes will be harder to detect (see
Alesina, 1989). Alt (1985) concluded that partisan effects are not permanent
but occur temporarily after a change of government.
In addition to the theoretical shortcomings of the early political business
cycle literature, the Nordhaus model also failed to attract strong empirical
support, with the econometric literature yielding inconclusive results (see
Mullineux et al., 1993). While McCallum (1978) rejected the implications of
the Nordhaus model for US data, Paldam (1979) could only find weak evidence
of a political business cycle in OECD countries. Later studies, such as those
conducted by Hibbs (1987), Alesina (1988, 1989), Alesina and Roubini (1992),
Alesina and Roubini with Cohen (1997) and Drazen (2000a, 2000b), also find
little evidence of a political business cycle in data on unemployment and GNP
growth for the US and other OECD economies. This also applies where the
timing of elections is endogenous (see Alesina et al., 1993). More favourable
results for the Nordhaus model are reported by Soh (1986), Nordhaus (1989),
Haynes and Stone (1990) and Tufte (1978), who found some evidence of preelection
manipulation of fiscal and monetary policy for the USA. In addition,
Drazen (2000a) distinguishes between empirical predictions that focus on policy
outcomes (inflation, unemployment, growth) and those that focus on policy
instruments (taxes, government expenditure interest rates) and concludes that
‘the evidence for opportunistic manipulation of macroeconomic policies is
stronger than for macroeconomic outcomes’. With respect to policy outcomes
there is more support for the opportunistic political business cycle theory
coming from post-electoral inflation behaviour than can be found in the preelectoral
movement of real GDP and unemployment. Drazen also concludes
that the evidence in favour of opportunistic manipulation of policy instruments
is much stronger for fiscal policy than it is for monetary policy.
A significant problem for partisan theories in general is the argument and
evidence presented by Easterly and Fischer (2001) that low inflation helps the
poor more than the rich. Inflation acts as a financial tax that hits the poor
disproportionately because they tend to hold more of their wealth in cash
relative to their income than the rich, whereas the rich are more likely to have
access to various financial instruments that allow them to hedge against
inflation. In addition the poor obviously depend more on minimum wages
and state-determined income payments that are not always indexed to protect
against the effects of inflation (see Snowdon, 2004b). Easterly and Fischer
present evidence drawn from an international poll of over 30 000 respondents
from 38 countries and the responses indicate that the poor themselves are
more strongly averse to inflation than those with higher incomes. This undermines
a key assumption of Hibbs’s model that the rich are more inflation-averse
than the poor.
By the mid-1980s the politico-economic literature had reached a new ebb.
Zarnowitz (1985), in his survey of business cycle research, devotes one
footnote to the idea of an electoral cycle and refers critically to the ‘strong’
and ‘questionable’ assumptions of such models as well as the lack of supporting
evidence. The same neglect is also a feature of Gordon’s (1986) edited
survey of The American Business Cycle.
Following a period of relative neglect, the literature on the relationship
between politics and the macroeconomy underwent a significant revival since
the mid-1980s (see Willet, 1988). Economists responded to the rational expectations
critique by producing a new generation of rational politico-economic
models. Like the first phase in the 1970s, the second phase of politicoeconomic
models consisted of opportunistic and partisan versions of the
interaction between politics and the macroeconomy. In the next two sections
we examine the main features of rational political business cycle models and
the rational partisan theory.
No comments:
Post a Comment