Conf 26 I
Conf 26 I
Conf 26 I
John T. Woolley*
Almost all observers agree that in some sense political factors matter
in the conduct of monetary policy. There is much less agreement as to which
political factors are most important, or as to exactly how those factors
influence monetary policy. In this essay, I will advance a rudimentary typol-
ogy of political factors that have been of interest to students of the conduct
of economic policy. I will then examine in some detail the consequences of
a subset of those factors for policy in three countries: Britain, France, and
Germany. I conclude with a broader discussion of the importance of politi-
cal factors and how they might be studied.
Io A Nudimentary Typology of Political Factors
The first distinction that I wish to draw is between factors that directly
involve the legally constituted process of governing and other factors, that,
while important for governing, are more accurately conceived as character-
istics of the nongovernmental organization of society. This is, of course, the
familiar distinction between "state" and "society." The second distinction is
between slowly and rapidly changing political factors. This, also familiar, is
the distinction between structural and variable factors. Where the line is
drawn between state and society is, as we shall see, extremely important,
and appears to have significant consequences for monetary policy. Politics
involves the use of authority to resolve distributional issues, and it is clear
that decisions about the range of societal affairs subjected to authoritative
(as opposed to market) decisions are fundamental political decisions. We
should not fail to recognize that monetary policy is both constrained by
those decisions and may be part of a continuing process of remaking or
refining those decisions.,
Illustrative examples of these four categories of political factors are
presented in Table 1. Each type is numbered moving clockwise from the
upper left quadrant. Most casual references to "politics" are references to
type I politics, or variable governmental politics. Because type I politics
figures prominently below, discussion here is quite brief. I distinguish three
major subcategories within type I politics. First, there is the politics in-
volved in determining the official ruling party or coalition. For Britain,
France, and Germany, this means partisan electoral competition. Second,
*Assistant Professor of Political Science, Washington University, St. Louis, Mo. Research
for this paper was supported in part by a grant from the Social Science Research Council of
New York. The paper has benefited from many conversations with James Alt. The author
gratefully acknowledges the research assistance of Donald Davison and Diane Schmidt.
177
178 MONETARY POLICY
Table 1
A Rudimentary Typology of Political Factors of Interest to Students of
Macroeconomic Policy
Governmental Nongovernmental
Type I Type II
Election Contests Wage Bargaining
Dynamics of Public Opinion Strike Behavior
Variable Legislative Politics Business Confidence
Bureaucratic Politics
Interest Group Politics
Type IV Type III
Division of Power between Degree of Unionization
Executive and Legislature Links of Parties and Unions
Structure and Control of Organization of Business Sector
Structural Public Bureaucracy Financial Structure
Central Bank Independence National Preferences for
Inflation/Unemployment
1In addition to this kind of ordinary interest group effort to influence policy decisions,
type I politics can incorporate another. In the U.S. context, at least, conflicts among contend-
ing groups of economists for influence over policy can be usefully conceived of as analogous to
interest group politics (Woolley, 1982).
POLITICAL FACTORS WOOLEY 179
protection.2
Type IV politics, or structural government policies, is, like type I poli-
tics, close to the commonsense notion of politics. This includes such consti-
tutional characteristics as the relative power of the head of government and
the legislative body, the rules governing electoral competition, the degree
of central bureaucratic control, and, of particular interest in the present
context, the formal relationship of central bank and government. Does
fluctuating support in parliament predictably lead to accommodation of
dissident views by the prime minister (chancellor, president)? Are members
of legislative bodies limited in their access to policymakers in central banks,
or, as in the United States, are they in the position to subject central bank-
ers to public hearings, to require frequent reports, and, potentially, to
change central bank law despite opposition of the head of government?
The linkages between these various kinds of politics and monetary
policy decisions are, of course, complicated. I suggest one possible set of
linkages in Figure 1. Whether, and how, type I politics affects monetary
Figure 1
Type I Politics
(mediated by (Indirect Pressure)
governmental structure)
Type g Politics
Type ~V Politics
*l’he data for Germany and Britain were provided in part by Thomas Yantek, and the
data for France in part by Michael Lewis-Beck. The Yantek and Lewis-Beck sets have been
extended to early 1982. The British data are Gallup poll data on voting intentions and are
published in the Gallup Poll Index. The French data are the IFOP series on approval pub-
lished regularly in France-Soir.
POLITICAL FACTORS WOOLEY 185
Table 2
Government Popularity and the Conduct of Monetary Policy
Coefficients on Popularity Variables Added to the Basic HR Model
POPDEF POPSUR
Bank of England
-0.160 -0.130
(-3.091) (-2.535)
Bank of France
Presidential Popularity 0.0925 -0.0233
(1.725) (-0.487)
Prime Ministerial Popularity 0.076 (-0.056)
(1.758) (-1.311)
Bundesbank
-0.0904 -0.194
(-0.781) (-1.726)
t-statistics in ().
186 MONETARY POLICY
Table 3
Preelection Conduct of Monetary Policy, Noncrisis Elections
(Sign of Coefficient on Preelection Dummy Variable Added to HR Model)
Bank of England Bank of France Bundesbank
Year Year Year
Parliament National Assembly Bundestag
66 - 0.273 67 - 0.085 72 - 1.542
( - 0.452) (- 0.130) ( - 2,235)
70 - 0.820 73 - 0.064 76 0.324
(- 1.291) (- 0.097) (0,385)
79 0.591 78 0.560
(0.825) (0.863)
Presidential
65 -0.111
(-0.170)
81 - 0.045
(-o.o51)
t-statistics in ( ).
cept for the German election of 1972, there is no indication that interest
rates in any of these countries were moved prior to elections in a fashion
significantly different than would have been expected given the responses
of officials throughout the entire estimating period. Only one of the two
SPD-called elections fits the political business cycle hypothesis, but given
the expectations outlined above and the extensive anecdotal evidence to
that effect that the Bundesbank is very independent, it is surprising that the
only significant preelection coefficient is found in the German case.
While most governments do not try, or are unable, to drive interest
rates down for political reasons prior to elections, there is much stronger
indication that decisive policy moves do follow elections as suggested by
the mandate hypothesis (see Table 4). The dummy variable used here takes
the value 1 in the four quarters after the election, else 0. The idea underly-
ing this definition is that the attempt to pursue a mandate is a brief one,
and that after four quarters more usual forces dominate monetary policy. In
Britain, the Labor governments of the mid-1970s reduced interest rates
significantly following both elections. In France each National Assembly
election since 1968 was ,followed by a substantial move in interest rates. In
the elections of 1968 and 1973, when the center-right coalition won comfort-
able victories, there immediately followed a more restrictive monetary poli-
cy. In 1978, however, the center-right barely won--mostly because of deep
divisions on the left--and that election was followed by a significant de-
crease in interest rates.
188 MONETARY POLICY
Table 4
Post-election Conduct of Monetary Policy All Elections
(Coefficient of Post-election Dummy Variable Added to HR Model)
Bank of England Bank of France Bundesbank
Year Year Year
Labor Victories National Assembly Bundestag
66 - 0.242 67 - 0.180 69 1.0335
( - 0.409) ( - 0.274) (1.055)
741 - 1.578 68 1.195 72 2.633
( - 2.537) (1.957) (3.557)
742 - 1.114 73 1.505 76 0.612
(- 1.693) (2.385) (0.778)
78 - 1.333
(-2.208)
Conservative Victories
70 - 0.440
(-0.746)
Presidential Elections
79 2.065
(3.133) 65 - 0.027
(-0.041)
69 - 0.3216
(-0.458)
74 0.9812
(1.354)
Conservative Incumbents Giscard Schmidt
0.389 - 3.325 - 0.781
(1.237) (3.454) ( - 0.929)
Pompidou
- 1.277
(-2.365)
t-test in ().
albeit ineffectually, predictably tries to find a whip for flogging the Fed
whenever interest rates are high (Woolley, 1984b).
characterization of the role of the Federal Reserve as being "to take away
the punch bowl just when the party really gets going."
Following Lindblom again, we should expect that between countries
the degree of attention to business confidence will vary with the degree of
public sector involvement in capital allocation and credit markets. The
greater the dependence on private markets, the greater the sensitivity to
business confidence. Thus, sensitivity should be greatest in Germany and
least in France, with Britain somewhere in between.
The data on business confidence in all three countries are drawn from
relatively lengthy time series of surveys of the outlook of entrepreneurs for
the next few months (often a calendar quarter). The survey results are
usually interpreted as revealing the degree of optimism of entrepreneurs.
By convention, the results are reported as the difference between the per-
cent responding positively and the percent responding negatively to ques-
tions about the future. Thus, these series cycle between some positive
number of around 40 when optimism is high and an equivalent negative
figure when pessimism is predominant.5 This is the measure of business
confidence used here, introduced with a one-quarter lag.
The basic confidence measure is labeled CONE The target confidence
level is defined as the mean of CONE and periods of surplus and deficit are
defined by CONF2, equal to CONF-MEAN(CONF). BUSDEF1, for per-
iods of confidence deficit, is equal to - I(CONF2) if CONF2 is less than or
equal to 0 (else = 0). BUSSUR1, for period of confidence surplus, is equal
to CONF2 if CONF2 is greater than 0.
The signs on the coefficients when both BUSSUR1 and BUSDEF1 are
entered in the basic HR model are, unlike the case of popularity, exactly as
expected in all three countries (see Table 5). Furthermore, the relative
sensitivity of policy to business confidence in the three.cases is also exactly
as expected. The most significant coefficient on BUSDEF1 occurs in the
German case--where POPDEF1 had no statistically significant coefficient.
The next most significant coefficient occurs in the British case but this
coefficient is less significant than the British coefficient on POPDEF1. In
France, the results are weak but have the correct sign, which was not the
case with POPDEF1. The Martin hypothesis fits in Britain and France, but
not in Germany, where confidence surplus produces no systematic
response.
5For Britain the data are the "optimism" series from the CBI Industrial Trends Survey
Data series. The series was provided directly by the CBI. Missing values are replaced with the
mean of adjacent observations. For France, data are quarterly averages of the monthly surveys.
in the "production prdvue" series from the INSEE enqu~te mensuelle aupr~s des chefs d’entre-
prises industrielles, "ensemble des branches." For the period since 1969, this series is pub-
lished in Tendences de la Conjoncture; earlier data are found in Etudes" et Conjoncture:
Suppldment. The German data are quarterly averages of monthly observations from the IFO
series "Geschaeftserwartungen fuer die naechsten 6 Monate: Investions.guter" and were pro-
vided by the Bundesbank.
192 MONETARY POLICY
Table 5
Business Confidence and the Conduct of Monetary Policy (Confidence Measures
Added to Basic HR Model)
BUSSUR1 BUSDEFt
(Confidence surplus) (Confidence deficit)
Bundesbank
0.0212 -0.0636
(0.729) ( - 2.305)
Bank of England
0.0228 - 0.0147
(2.518) (- 1.746)
Bank of France
0.0313 - 0,0209
(1.253) (-1.127)
t-statistics in ().
This brief excursion into the implications of "group politics" for mone-
tary policy provides support for the notion that monetary politics is not
interest-group politics. Monetary policy does attend to the needs of busi-
ness, but the degree to which this is true depends on the role of the state in
society.
6It is certainly true that the popularity and business confidence variables are not strictly
speaking exogenous. It is not my contention that these quantities are unrelated to the econom-
ic variables included. It is clearly the case, however, that the included economic variables do
not do a particularly good job of accounting for variation in either popularity or business
confidence. It is possible then that both variables convey additional valuable information
about the condition of the political economy to policymakers, and the models estimated here
suggest that this is the case.
POLITICAL FACTORS WOOLEY 193
Table 6
Full Models Incorporating All Relevant Political Variables
Bank of England (65Q2 - 80Q4)
BRQ = 0.482 + 0.00031 DOBA - 0.0034 DOFRES + 0.775 BRQ1
(0.877) (2.469) ( - 3.937) (10.371 )
+0.197 REDQ1 - 0.758 DCPI + 0.00032 PSBRQ
(3.080) ( - 0.234) (2.924)
-0.128 POPDEF1 - 0.0335 POPSUR1 - 0.0162 BUSDEF1
(-2.636) ( - 0,675) (-2,118)
+0.0222 BUSSUR1 + 1.421 POST79
(2.574) (2.368)
~~ = .91 DW = 2.053
Bank of France* (64Q2 - 80Q4)
IBRQ = 1.388 = 12.789 DFRDMXRQ + 60.185 DCPI
(3.227) (6.246) (4.308)
+0.289 REDQ1 + 0.012 GDEF1 + 1.184 DIPIC70
(4.603) (1.533) (1.515)
+1.039 POST68 + 1.395 POST73 - 1.156 POST78
(1.813) (2.483) ( - 2.055)
-0.889 POMPIDOU - 1.338 GISCARD
( - 1.684) ( - 1.261 )
~2 = .86
t-statistics in ().
*Models corrected for first-order autocorrelation by two-step full transform method,
The model for Germany includes the political cycle dummy variables
for the 1972 elections and the measures of business confidence. In the full
model, the "mandate" variable, POST72, proves to be insignificant, leav-
ing only two significant political variables, one for the PRE72 election
period, and one for the response to business confidence deficits. These
results are, overall, a strong confirmation of Bundesbank independence
from type I politics. It is also a striking confirmation of the importance of
business confidence in the conduct of monetary policy in market-dominat-
ed systems, exactly as Lindblom’s hypothesis would lead us to expect. This
type II political effect is, of course, fundamental to stability and growth in
this kind of political economy.
The British model differs from the German principally by including
popularity variables. With the exception of the POST79 variable (marking
the onset of the Thatcher regime) no "mandate" variables are significant
when the popularity measures are included. The coefficient on the popular-
194 MONETARY POLICY
much less easily obtained. We would need to know much more, in some
detail, about the consequences for various sectors and groups of changes in
both the policy instrument and the policy targets. We would need an elabo-
rate and politically sensitive structural model of the political economy. The
forecasting models in use today are for the most part inappropriate for this
kind of task. The objective of such an ideal model would be to reveal in
detail the various implications of the policy actions that policymakers take.
We could than know the consequences for particular groups or economic
sectors if the central bank moved its interest rate a certain amount. We
would also know exactly how far that action went toward stabilizing the
many targets included in the reaction function. One could objectively ana-
lyze the political meaning of the entire reaction function because one could
say clearly exactly what kinds of consequences were being avoided by
policymakers.
Such a model does not exist and would probably prove impossible to
construct. However, it is still possible that portions of such a model could
be developed, at least qualitatively. It is also possible that central bankers
and others involved with monetary policy might be persuaded to reveal
more about their own judgments as to the possible consequences of follow-
ing a very different policy from the ones they have selected. If we are to
move further toward understanding the political and social antecedents of
monetary policy, we cannot avoid the need to advance explicit interpreta-
tions of the ways policy routines of the sort modeled in reaction functions
reflect and reinforce an existing political solution for society.
In such research, one should be prepared for the possibility that mone-
tary authorities may not try to anticipate in any detail the consequences of
their actions that are more than a step or two removed from the actions
themselves. One should also be prepared for the possibility that monetary
authorities, rather like "satisficing" actors described by Herbert Simon,
ordinarily consider a range of possible policy actions that is relatively nar-
row and bounded by rough rules of thumb (e.g., "if we let interest rates
move up that much we’d have chaos in the markets"). Given the nature of
things, such rules are largely nonfalsifiable. Such rules make it unnecessary
to attempt to project consequences in very much detail; such projections
would in any case be of dubious value since the alternatives of interest are,
more likely than not, well outside the range of previous experience.
In short, there is no reason a priori why we should expect policy-
makers themselves to be able to articulate, even in conditions congenial to
frank reflection, the political formulae that underly their behaviors. What
we can predict, I suspect with very high levels of certainty, is that central
bankers rarely propose dramatic changes in monetary policy; the reasons
for this are as much political as economic: uncharted economic ground is
strewn with political hazards. The fact that this conservativism is usually
not recognized as reinforcing a political formula should not obscure its
political nature for outside analysts.
196 MONETARY POLICY
~,e~erences
Alt, James E. and Alec Chrystal. 1983. Political Economics. Berkeley: University of Califor-
nia Press.
Air, James E. and John T. Woolley. 1982. "Reaction Functions, Optimization, and Politics:
Modelling the Political Economy of Macroeconomic Policy," American Journal of Politi-
cal Science (November): 709-740.
Aronson, Jonathan D. 1977. Money and Power: Banks and the World Monetary System.
Beverly Hills: Sage, 1977.
Frey, B. and Schneider; E 1978a. "An Empirical Study of Politico-Economic Interaction in the
United States," Review of Economics and Statistics 60: 174-183.
___. 1978b. "A Politico-Economic Model of the U.K. ," Economic Journal 88: 243-253.
1979. "An Econometric Model with an Endogenous Government Sector," Public
Choice 34: 29-44.
Gordon, Robert J. 1977, "World Inflation and Monetary Accommodation in Eight Coun-
tries," Brookings Papers" on Economic Activity #2: 409-468.
Heclo, Hugh. 1978. "Issue Networks and the Executive Establishment," in Anthony King, ed.
The New American Political System, Washington, D.C.: The American Enterprise
Institute.
Hibbs, Douglas A., Jr. 1977a. "Political Parties and Macroeconomic Policy," American Politi-
cal Science Review 71: 1464-1487.
__. 1977b. Trade Union Power, Labor Militancy and Wage Inflation; A Comparative Anal-
ysis. Cambridge, Mass.: Center for International Studies, Massachusetts Institute of
Technology.
Hodgman, Donald R. and Robert W. Resek. 1983. "Determinants of Monetary Policy in
France, The Federal Republic of Germany, Italy, and the United Kingdom: A Compara-
tive Analysis." This volume.
Katzenstein, Peter J., ed. 1977. Between Power and Plenty: Foreign Economic Policies of
Advanced Industrial States. Madison: University of Wisconsin.
Laidler, David. 1976. "Inflation--Alternative Explanations and Policies: Tests on Data
Drawn from Six Countries" in K. Brunner and A. H. Meltzer, eds. Institutions, Policies
and Economic Performance. Amsterdam: North Holland.
Lindblom, Charles E. 1977. Politics and Markets: The WorM’s Political-Economic Systems.
New York: Basic Books.
Nordhaus, W. 1975. "The Political Business Cycle," Review of Economic Studies’ 42: 169-190.
Parkin, M. and R. Bade. 1978. "Central Bank Laws and Monetary Policies: A Preliminary
Investigation." Mimeographed. London, Ontario: University of Western Ontario.
Sachs, Jeffrey D. 1979. "Wages Profits and Macroeconomic Adjustment: A Comparative
Study," Brookings Papers on Economic Activity, #2.
Schmidt, Manfred G. 1982. "Does Corporatism Matter? Economic Crisis, Politics and Rates
of Unemployment in Capitalist Democracies in the 1970s," in G. Lehmbruch and E C.
Schmitter, eds. Patterns of Corporatist Policy-Making. Beverly Hills: Sage.
Suleiman, Ezra N. 1974. Politics Power and Bureaucracy in France: The Administrative Elite.
Princeton: Princeton University Press.
__. 1978. Elites in French Society: The Politics of Survival. Princeton: Princeton University
Press.
__. 1980. "Presidential Government in France," in Richard Rose and Ezra N. Suleiman,
eds. Presidents and Prime Ministers. Washington, D.C.: The Amerian Enterprise
Institute.
Tufte, Edward R. 1978. Political Control of the Economy. Princeton, N.J.: Princeton Universi-
ty Press.
Woolley, John T. 1977. "Monetary Policy Instrumentation and the Relationships of Central
Banks and Governments," The Annals of the American Academy of Political and Social
Sciences 434 (November): 151-173.
POLITICAL FACTORS WOOLEY 197
__. 1982. "Monetarists and the Politics of Monetary Policy," The Anna~ of the Academy of
Political and Social Sciences 459 (January): 148-160.
__. 1984a. "Central Banks and Inflation: Influence and Independence," in Leon N. Lind-
berg and Charles Maier, eds. The Politics and Sociology of Global b~flation and Reces-
sion. Washington, D.C.: The Brookings Institution.
__. 1984b. Monetary Politics: The Federal Reserve and the Politics ~f Monetary Policy. New
York: Cambridge University Press.
Appendix A.
All British Models
Dependent Variable = BRQ (65Q2 - 80Q4)
Inter-
cept BRQ1 DOBA DOFRES REDQ1 DCPI PSBRQ POLITICAL VARIABLES R2 DW
POPSUR1 POPDEF1
1. 1.043 0,785 0.00033 - 0.00025 0,223 - 5.337 0.00032 - 0.130 - 0,160 .88 1.832
(2.088) (9.976) (2.355) (-2.781) (3,108) (-1,518) (2.529) (-2.535) (-3.091)
PRE66
2. 0,556 0.746 0.00024 - 0.00027 0.217 - 2.430 0.00037 - 0.273 .86 1.790
(1.063) (8.988) (1.611 ) ( - 2,838) (2.741) ( - 0.654) (2.783) ( - 0.453)
PRE70
3. 0,457 0.730 0.00021 - 0.00027 0.257 - 2.383 0.00034 - 0.820 .88 1.812
(0.934) (8.879) (1.413) ( - 2,886) (3,200) ( - 0.657) (2,529) ( - 1.291)
PRE79
4. 0.463 0,745 0.00024 - 0.00024 0.218 - 1.695 0.00034 0.591 .86 1.806
(0.937) (9.048) (1.657) ( - 2.488) (2,815) ( - 0.459) (2,482) (0,825)
POST66
5. 0.534 0.742 0.00024 - 0.00027 0.233 - 2.436 0.00038 - 0.242 .86 1.786
(1.042) (8.975) (1.614) (-2.856) (2,886) (- 0.653) (2,799) (- 0.409)
POST741
6. 0,108 0,714 0,00036 - 0,00032 0,298 - 0,459 0.00036 - 1.578 .89 1.985
(0.219) (9.020) (2.432) ( - 3.473) (3.783) ( - 0.129) (2,841) ( - 2,537)
POST742
7, 0,385 0.703 0.00030 - 0.00031 0,255 0,581 0.00039 - 1,114 ,88 1,904 ©
(0.789) (8.382) (2.000) (-3.276) (3.287) (0.148) (2.979) (- 1.693)
ALL CONSERVATIVE
8. 0.473 0,747 0.00028 - 0.00026 0.205 - 2.414 0.00038 0,389 .88 1,186
(0.965) (9,140) (1,869) (-2.931) (2.632) (-0.665) (2.857) (1.237)
POST70
9. 0.542 0.736 0.00022 - 0.00027 0.228 - 2.003 0.000364 - 0.440 .86 1.798
(1,081) (8.885) (1.463) ( - 2.852) (2.952) ( - 0.547) (2.708) ( - 0.746)
POST79
0.667 0.00032 - 0.00039 0.172 0.485 0.00034 2.065 .89 2.070
10. 1.240
(2.395) (8.332) (2.294) (-4.099) (2.350) (0.139) (2.722) (3.133)
BUSSUR1 BUSDEF1
0.00020 - 0.00029 0.241 - 0.9334 0.00041 0.0228 - 0.0147 .90 1.793
11. - 0.194 0.779 (- 1.746)
( - 0.396) (10.330) (1.474) ( - 3,339) (3.419) ( - 0.274) (3.341) (2.518)
MTGPRESS
0.00011 -0.00016 0.195 -6.087 0.00032 0.244 .88 1.766
12. 0.398 0.788 (2.271)
(0.836) (9.648) (0.752) ( - 1.577) (2.599) ( - 1.554) (2.488)
0.742 0.00024 - 0.00027 0.225 - 2.161 0.00037 .87 1.782
13. 0,481
(0.977) (9.047) (1.611 ) ( - 2.873) (2.926) ( - 0.593) (2.814)
t-statistics in ( )
2OO MONETARY POLICY
Bank of England
BRQ = Bank Rate
BRQ1 = Bank Rate lagged one quarter
DOBA = Change over prior quarter in official borrowing abroad, lagged one quarter
DOFRES = Change over prior quarter in official reserves lagged one quarter
DCPI = Percentage change over four quarters in consumer price index, lagged one quarter
(ln(CPIt.I/CPIt_5))
PSBR = Public sector borrowing~ requirement, current quarter
Bundesbank
LR3MOSQ = Three month money-market loan rate
DCPI76 = Percentage change ow;r four quarters in consumer price index, lagged one quarter
(base 1976) (ln(CPIt_I/CPIt.5))
MDLSPRQ5 = Percentage change over four quarters in DM/$ exchange rate when seasonally
adjusted current account balance of payments is greater than 0.
MDLSPRQ6 = Same as MDLSPRQ5 when current account balance of payments is less than
or equal to 0.
CAPUTB1 = Rate of capacity utilization lagged one quarter.
REDQ1 = Eurodollar rate lagged one quarter.
Bank of France
IBRQ = Quarterly average for overnight money market rate
DFRDMXRQ = Percentage change over four quarters in the spot exchange rate, French
Franc/DM.
DCPI = Percentage change over four quarters in the consumer price index, lagged one quarter
(ln(CPIt_,/CPIt.s))
REDQ1 = Eurodollar rate lagged one quarter
GDEF1 = Government budget deficit lagged one quarter
DIPIC70 = Percentage change over four quarters in industrial production index (base 1970),
lagged one quarte~:
Appendix B: All German Models*
Dependent Variable = LR3MOSQ (69Q3 - 80Q1)
Inter- ~2
cept DCPI76 MDLSPRQ5 MDLSPRQ6 CAPUTB1 REDQ1 POLITICAL VARIABLES
POPDEF1 POPSUR1
1. - 36.814 93.646 - 5.850 - 6.051 0.442 0.288 - 0.090 - 0.194 .75
( - 6.144) (5.825) ( - 2.077) ( - 1.041 ) (5,877) (2.815) ( - 0.781 ) ( - 1.726)
PRE 72
2. - 36.065 99.549 - 6.155 - 6.593 0.424 0.305 - 1.547 .82
( - 8.931 ) (7.243) - 2.397) ( - 1.115) (8.687) (3.547) ( - 2.235)
PRE 76
3, -34,112 91.512 - 6.080 - 5.403 0.396 0.379 0.324 .73
(-6.761) (5.626) ( - 1.970) ( - 0.890) (6.623) (4.119) (0.385)
POST 69
4. -30.509 99.146 - 5.534 - 4.286 0.350 0,362 1.033 .71
(5,640) (5.260) ( - 1,960) ( - 0.748) (5.271) (3.819) (1.055)
POST 72
5. -31.545 72.157 - 1.375 - 3.005 0.375 0.403 2.633 .87
( - 8.637) (5.407) ( - 0.521 ) ( - 0.529) (8.725) (5.726) (3.557)
POST 76
6. - 33.472 92.860 - 5.274 - 3.950 0.387 0.393 0.612 .71
( - 6.438). (5.350) - 1.848) ( - 0.675) (6.292) (4,036) (0.778)
BUSSUR1 BUSDEF1
7. - 33.924 109.581 - 6.640 - 4.494 0.391 0.358 0,021 - 0.064 .83
(-8,577) (7.717) - 2,681 ) ( - 0.805) (8.159) (4.689) (0.729) ( - 2.305)
SCHMIDT
8. - 26.084 93.443 - 6.630 - 6.347 0.299 0.430 - 0.781 .73
(-3.259) (5.255) - 2.486) ( - 1,253) (3.230) (4.124) ( - 0.929)
9. -31.558 89.412 - 5.368 - 4.612 0.368 0.374 .68
( - 5,921 ) (4.946) ( - 1.881 ) ( - 0.820) (5.771) (3.842)
t-statistics in ( ).
*All models corrected for first-order autocorrelation by two-step full transform method, to
Appendix C: Al! French Models1 Po
Dependent Variable = IBRQ (69Q2 - 80Q4)
Inter-
cept DFRDMXRQ DCPI REDQ1 GDEF1 DIPIC70 POLITICAL VARIABLES ~2
1. 1.980 9.104 41.628 0.315 0.012 1.071 .66
(3,422) (3.903) (5.378) (4.018) (1.610) (1.472)
PPDEFI* PPSUR1
2. 2.333 8.448 34.480 0.315 0.011 1.018 0.093 - 0.023 .67
(3.749) (3.616) (4.080) (4.107) (1.468) (1.396) (1.725) ( - 0.487)
PMDEF1 ** PMSUR1
3. 2.632 9.450 33.361 0.310 0.014 0.801 0.076 - 0.056 .65
(3.680) (3.914) (3.888) (3.786) (1.727) (0.940) (1.758) (- 1.311)
PRE65
4. 2.005 9.080 41.413 0.314 0.012 1.071 - 0.111 .65
(3.355) (3.854) (5,235) (3.981) (1.595) (1.460) ( - 0.170)
PRE67
5. 1.992 9.O9O 41,404 0.316 0.012 1.070 - 0.085 .65
(3.390) (3.386) (5.180) (3.975) (1.595) (1.458) ( - 0.130)
PRE73
6. 1,982 9.140 41.693 0.314 0.012 1.080 - 0.064 .65
(3.374) (3.862) (5.368) (3.916) (1.595) (1.461) (-0.097)
PRE78
7. 1.935 8.910 40.584 0.326 0.011 1.052 0.560 .65
(3.364) (3.791) (5.194) (4.111 ) (1.538) (1.438) (0.863)
PRE81
8. 1.954 9.186 41.690 0.317 0.012 1.069 -0.045 .66 ~
(3.205) (3.872) (5.346) (3.847) (1.496) (1.426) ( - 0.051 )
~O
POST67 ~1
9. 2.01 9.066 41.281 0.314 0.012 1.079 - 0.180 .65 ~
(3.387) (3.846) (5.231) (3.987) (1.600) (1.469) (-0.274)
POST68 ~
10. 1.716 10.418 43.006 0.321 0.012 1.105 1.195 .73 ~
(3.369) (4.690) (6.269) (4.359) (1.603) (1.464) (1.957) ~O
POST73
11. 1.980 8.890 45.282 0.271 0.012 1.120 1.505 .68
(3.577) (3.955) (5.956) (3.495) (1.712) (1.596) (2.385)
POST78
12. 1.825 9.572 41.933 0.340 0.012 1.071 - 1.333 .73
(3.655) (4.359) (6.207) (4.644) (1.569) (1.428) ( - 2.208)
POST65
13. 1.985 9.099 41.562 0.315 0.012 1.071 - 0.027 .65
(3.351) (3.860) (5.214) (3,980) (1.595) (1.459) ( - 0.041 )
POST69
14. 1.958 9.593 41.093 0.322 0.0t 2 1.078 - 0.322 .65
(3.407) (3.808) (&258) (4.030) (1.587) (1.466) ( - 0.458)
POST74
15. 2.213 10,071 37.968 0.305 0.011 1.038 0.981 .65
(3.687) (4.115) (4.618) (3.904) (1.504) (1.443) (1.354)
BUSSUR1 BUSDEF1
16. 1.664 8.346 48.460 0.288 0.012 1.246 0.031 - 0.021 .67
(2.794) (3.552) (5.668) (3.636) (1.670) (1.695) (1.253) (-1.127)
POMPIDOU GISCARD
17. 1.058 12~890 83.990 0.257 0.012 1.175 - 1.277 - 3.375 .77
(2.075) (5.745) (6.087) (3.548) (1.628) (1.581) ( - 2.365) (-3.451)
t-statistics in ( ).
I-All models corrected for first-order autocorreiation by two-step full transform method.
*Presidential Popularity
*Prime Ministerial Popularity
Discussion
Marcello de Cecco
204
DISCUSSION DECECCO 205