Data Appendix for Currency Politics: The Political Economy of Exchange Rate Policy (Princeton University Press, 2015)

Data Appendix for Currency Politics: The Political Economy of Exchange Rate Policy (Princeton University Press, 2015)

Table of Contents

• 1. US 1800s
  • 1.1. Download
  • 1.2. Online Documentation
• 2. Latin America
  • 2.1. Download
  • 2.2. Online Documentation
• 3. EU
  • 3.1. Download
  • 3.2. Online documentation

1 US 1800s

1.1 Download

• US 1800s - All datasets and documentation (ZIP, 3.5M)

1.2 Online Documentation

1.2.1 Overview

The final datasets are generated into data.files.YYYYMMDD in the root directory. If not using the shell scripts, then the *.dta files will be found in data/dataFiles/processed/for_analysis, and the user scripts *.do can be copied over manually from the otherscripts directory (note that they must be in the same directory as the datasets to work).

The current naming convention for the House of Representatives data is:

h##.[description of topic issue].[pooled|wide].[scd|sl].dta

The fields are:

h##

gives the Congress number; if there are two, then they are both listed at the beginning of the filname.

[description of topic issue]

This is a textual description of the topic issue that is included in the dataset. Rather than analyzing all votes at a go, it makes more sense to group them by votes that are related to particular topic issues, or large acts, in order to provide consistency in what a "hard money" vote means. More info on the topic issues is give below.

pooled|wide

indicates the shape of the datafile. Wide datasets give one observation per Congressional district. Votes for separate motions appear as different columns, which are named v##c, where ## gives the vote number, which is unique to each Congress.

Pooled datasets have one observation per vote, i.e., each Congressional district appears as many times as there are votes included in the issue topic. This is potentially better for providing an overall picture, but standard errors will need to be corrected for correlations between observations. The most standard way to do this is to cluster by Congressional district.

scd|sl

Indicates the level of the economic data of the dataset. sl indicates state-level data, so all Congressional districts in the same state will share the same economic data. scd indicates Congressional district level data.

For the Senate, a dataset has been generated to include the four votes analyzed in the original JEH paper. The naming convention is analogous, except that intial portion is s## to indicate Senate votes, and all the data is state level, so is indicated by s.

1.2.2 Topic issues

Votes are grouped according to topic issues. Since most votes cannot be completely described by the hard/soft vote dichotomy, pooling over all the votes means pooling values of the vote variable that might not be entirely consistent with each other. To help mitigate potential problems associated with this, we analyze groups of votes together; for instance inflation.bill datasets contain only the votes related to the Inflation Bill, and the vote variable will tell whether a Congressional district's representative votes for or against the Inflation Bill.

The file vote_codings.txt is dynamically generated from the comments of the associated R script, poolvotes.R, and will contain more up to date descriptions of the particular groupings, votes, and codings in the generated datasets. The most recent groups are described in the following section.

1.2.2.1 Topic definitions

This section describes the topics currently defined, and the direction that each vote is coded within each topic. Topic descriptions come from Jasmina Beganovic's notes on the secondary literature (see Additional Documentation).

Contraction Act

New Treasury Secretary High McCulloh asks Congress for authority to retire greenbacks as a first step toward resumption in December, 1856 and Congress approves by a vote of 144 to 6 pledging cooperative action in a contraction of the currency with a view to a return to specie (Unger, p.41). The original bill (Feb. 21, 1866) gives the Secretary a completely free hand in withdrawing greenbacks (V128). However, this bill is voted down and a more moderate bill is passed (V139) which allows the Treasury to retire from circulation up to $10 million of notes in the first 6 moths and no more than $4 million per month thereafter.

Table 1: Contraction Act

Suspension of the Contraction Act

However, during 1866-1867, McCulloh’s tight money supply policy was becoming increasingly unpopular among a heterogeneous group of businessmen, agrarians, and political opponents in part due to a business recession. Resistance culminated with a repeal of the 1866 Act passed by the House in December 1867. Senate attempted unsuccessfully to moderate the bill via amendments but passed it in February 1868(Unger, p.43).

Table 2: Suspension of Contraction Act

Bond repayment and Public Credit Act

The victory of the soft money forces on the question of money supply/contraction was followed by their defeat on the related issue of bond repayment (Unger, p.43). Western Democrats wanted the 5-20 bonds to be redeemed in greenbacks. Sherman, the Chairman of the Finance Committee came up with a funding bill in 1886 which allowed conversion of federal bonds into greenbacks in order to satisfy some pro-greenback constituents in his home state Ohio (Unger, p.88). Conservative Republicans took offense and stripped the bill of the offending features via amendments and passed it as a simple refunding measure (which did not become law due to President Johnson’s pocket veto). In the lead-up to the 1868 elections, Republicans took a strong stand against all forms of repudiation of the debt. The Republican victory in the 1868 Presidential elections gave the hard money camp strength and in March 1869 Congress passed the Pubic Credit Act, which calls for redemption of the 5-20s in gold.

Table 3: Bond repayment and Public Credit

Inflation Bill

In the fall of 1973 the country saw the end of the railroad boom and a panic which started a six year long economic recession. In late 1873, committees in both the House and Senate were working on bills to address the economic problems and respond to the expansionist pressures that had revived since the panic (Unger, p. 215). The bill introduced by the House (HR1572) authorized a modest expansion, a fund for currency elasticity, specie payments, and free banking. The Senate bill (S617) was a “…hodgepodge of expansionist, resumptionist and free banking elements that pleased no one entirely” (Unger, p. 217). Both bills were compromises, “designed to be all things to all men … with a jumble of elements pointing simultaneously in different directions” (Unger, p. 219). However, due to the general economic hardship, inflationist elements prevailed and both bills passed with amendments to make them more expansionary. In the Senate, S617 (which later became known as the Inflation Bill) was quickly amended into an inflationary measure (see V87, V88). Over time, the entire bill vas substituted and passed in its revised form. The House passed its own free banking bill and also approved the Senate measure by a decisive majority of 140 to 102 (Unger, p. 234).

The passing of the Inflation Bill was a product of a coalition of western and Pennsylvania Republicans. It was strongly opposed by conservative North-easterners from both parties with the support of merchants, bankers and reformers. Although it was generally believed the Pres. Grant favored the bill, he vetoed it in an unexpected move on April 22 (Unger, p.244). Senate voted on a motion to override the veto but fell short of the required two-thirds.

Table 4: Inflation Bill

Repeal of the Resumption Act

A compromise bill drafted by Senate Republicans that satisfied various Republican constituents by providing for free banking with a reduction of greenbacks and resumption by accumulation of a gold reserve, and resumption to gold at a certain date four years later (January 1, 1879). The bill was thought to be a showing of unity by Republican senators in an election year. It was passed in the Senate strictly along party lines and then in the House by a lame duck Congress with by moderate Republicans with the Democrats and Republican extremists opposed (Unger, p.259). It brought the country eventually to the gold standard but at the time of passing it was not taken very seriously due to its vagueness and lack of operative features.

Following the Resumption Act, several attempts were made in the House to repeal the unpopular measure, mainly led by liberal democrats. However, they encountered resistance from most Republicans and a group of North-eastern and Western Democrats. This split in the party proved challenging for the Democrats. Moreover, House rules on voting outside the regular committee time (2/3 majority required for victory) made the repeal more difficult.

Table 5: Repeal of the Resumption Act

Bland-Allison Act

Bland introduced the silverite bill in the House in November 1878. Senate moderates (including Alison) were able to soften the bill by removing the free coinage provision (Unger, p.362). The bill once more created a silver dollar, albeit at a limited coinage. It required the Treasury to buy between $2 and $4 million of silver per month at the fixed price of $1.29 per ounce to be coined into standard silver dollars1 (Timberlake, p.28). The bill was pro silver but a weak compromise. It was supported by the Greenbackers (Hicks, p.305), by most leading members of the Democratic party, and a significant number of Republicans from western silver-producing states. The executive branch was generally anti-silver in this period and when the bill came to President Hayes, he vetoed it. (Timberlake, p.28) Both the Senate and the House re-passed the bill with the required 2/3 majority.

Table 6: Bland-Allison Act

Sherman Act

Continued the principle of the Bland-Allison Act, but increased the government’s required purchase amount to 54 million ounces / year, which was roughly equal to the silver output of the U.S. Nevertheless, the price of silver continued to fall. The motivation for passing a pro-silver compromise on the Republican side stemmed from a promise they had made to certain constituents during the 1888 campaign in which Benjamin Harrison defeated Grover Cleveland in a close election. (Hicks, p.306)

Table 7: Sherman Act

1.2.3 Variable descriptions

1.2.3.1 Vote variables

1.2.3.2 Congressional district and state variables

All years

The following variables are found in both the 1870 and 1890 censuses, and so they appear in all datasets.

Table 8: Congressional district level variables available in all years

1890s

The following variables appear only in the 1890 census, and hence they are found only in datasets generated with votes for the later Congresses.

Table 9: Congressional district level variables available for 1890 only

1870s

The following variables are available in 1870s data only, at both the state and Congressional district level.

Table 10: Congressional district level variables available for 1870s only

The following variables are available only at the state level.

Table 11: State level variables available for 1870s only

2 Latin America

2.1 Download

• Latin America - Datasets (Annual and Monthly series) and documentation (ZIP, 3.7M)

2.2 Online Documentation

2.2.1 Annual

2.2.2 Monthly series

3 EU

3.1 Download

• Dataset and documentation (ZIP, 146K)

3.2 Online documentation

3.2.1 Note

All data updated through 1998 unless otherwise noted.

3.2.2 Dependent Variables

Coefficient of Variation (COV)

coefficient of variation over

12-month period Jan-Dec exchange rates (currency vs DM).  Calculated as standard deviation/mean, from IMF, International Financial Statistics Yearbook, various years.  COV2 is COV*100.

Depreciation (DEPREC)

annual depreciation of country currency vs

DM.  Calculated as (XRT – 1)*100 (see below).

3.2.3 Macroeconomic Variables

Lagged Growth Rate of GDP (GDPGRLAG)

Growth rate of GDP, lagged one

year. Data for 1971-1979 figures from Economic Survey of Europe, 1984-1985;  for 1980-1993 from OECD Historical Studies: 1960-1993.  Updated with OECD Economic Outlook, 1999.

Lagged Unemployment Rate (UNEMPLAG) - Percentage of the labor force unemployed, lagged one year (Standardized Rate used except for Greece).  Data taken from OECD Main Economic Indicators, Historical Studies:Prices, Labor and Wages 1962-1991, OECD Economic Outlook 1995 (volume 58) and 1999 (volume 62), OECD Main Economic Indicators, Historical Studies: 1960-1979, and Economic Survey of Europe, 1984-1985.

Lagged Current Account as a Percent of GDP (CURACLAG)

Current

account balance as a percentage of GDP, lagged one year. Data from OECD Economic Outlook, various years.

Difference in Terms of Trade Relative to Germany (DIFFTOFT)

Percentage point change in the terms of trade over the previous year, relative to Germany's.  An increase in this figure signifies an improvement in Germany's terms of trade relative to the country in question. Data from IMF International Financial Statistics Yearbook, various years.

NOTE: Updated with OECD data (percent change in export prices – percent change in import prices).  The original description above says that a positive number indicates an increase in Germany's TOFT, but the reverse data (i.e., country X's chgTOFT – Germany's chgTOFT) matches the old data most closely.  However, discrepancies do exist, so see the shaded columns in the dataset, which show my new calculations.  I have left these as separate columns, rather than integrate them into the original DIFFTOFT variable.

Industrial Correlation with Germany (INDCOR)

Correlation coefficient comparing the percent contribution to GDP of each ISIC 1-digit category and 2-digit categories for manufacturing  (ISIC code 3).  Because industrial structure changes slowly, the correlation coefficient is calculated for 1970, 1980, and 1990 only. Data from the OECD's Industrial Structure Statistics, various years.  Where data were missing from the OECD statistics, data were taken from the UN Yearbook of Industrial Statistics, various years.

NOTE: Since these data change only each decade, data were held constant for 1995-1998 at 1994 levels.

Change in Consumer Prices (CPI)  (%)

from IMF, /International

Financial Statistics Yearbook/, various years.

Terms of Trade (TOFT)

% change in export prices - % change in import prices, over the previous year.  From IMF, International Financial Statistics.  Some country data unavailable.

Trade correlation (TRACOR)

correlation coefficient computed using exports in the country by sector (see INDCOR) and comparing with Germany's figure.

NOTE: As with INDCOR, 1994 numbers held constant for 1995-1998.

Exchange rate against the DM (XRT)

End-of-year exchange rate of currency vs DM at time t (12/31/t), divided by exchange rate at time t-1.  From IMF, International Financial Statistics Yearbook, various years.

Exchange rate against the Dollar (EXCHRATE)

average exchange rate of currency vs $ at time t, when t denotes the current year. 

Exports to the EC as percent of GDP (ECX)

data from IMF, Direction of Trade Statistics, various years.

Exports to the DM Zone as percent of GDP (DMX)

data from IMF, Direction of Trade Statistics, various years.

Deposit Restrictions (DEPRES)

dummy for deposit restrictions, from Rose and IMF, Exchange Arrangements, various years.

Bilateral Payments with non-members (BIPAYNM)

dummy variable (1=yes), from Rose and IMF, Exchange Arrangements, various years.

Bilateral Payments with members (BIPAYM)

dummy variable (1=yes), from Rose and IMF, Exchange Arrangements, various years.

Capital Restrictions (CAPRES)

dummy variable (1=yes), from Rose and IMF, Exchange Arrangements, various years.

Capital Controls (CAPCTRL)

sum of the four previous variables, an index of capital controls (1-4, with 4 highest), from Rose and IMF, Exchange Arrangements, various years.

Trade balance (TRADE_B1)

Trade balance in millions of dollars.  Data from IMF,  International Financial Statistics Yearbook, various years.

Trade balance as % of GDP (TBALGDP)

Data from IMF, International Financial Statistics Yearbook, various years.

Change in trade balance as % of GDP (TBALCHG)

Change in the trade balance from the previous year, in percentage terms. Constructed from data for trade balance and GDP in IMF, International Financial Statistics Yearbook, various years.  Calculated as trade balance at time t – trade balance at time t-1, as percentage of GDP.

GDP

GDP in billions of local currency.

GDP_BIL

GDP in billions of dollars.

GDP_MIL

GDP in millions of dollars.

Government Deficit as % of GDP (DEFGDP)

Data updated from OECD Economic Outlook, general government financial balances, 1999.

3.2.4 Political Variables

Cabinet Center of Gravity (CABCG)

Party composition of the cabinet, weighted by ideological scores using a scale constructed by Geoffrey Garrett.   Data through 1991 provided by Geoff Garrett; updated using European Journal of Political Research Political Data Yearbooks, various years. Parties classified using codings by Castles and Mairs (1984, EJPR 12: 73-88). /NOTE: 1=far left, 2=social democratic, 3=Christian Democratic, 4=mainstream conservative, 5=far right./

Election

Number of parliamentary elections per year (usually 1 or 0).  Data obtained from Mackie, Thomas T. and Richard Rose International Almanac of Electoral History (Washington D.C.: Congressional Quarterly, 1991), National Elections (various years), and the European Journal of Political Research (various years).**

% of Seats Held by Government Parties

Percentage of legislative seats won by the government parties in the election at time t, where t denotes the current observation (SEATT) and at time t-1 (SEATT1).  Constructed in G. Bingham Powell, Jr. and Guy D. Whitten, “ A Cross-National Analysis of Economic Voting,” American Journal of Political Science 37(2): 391-414, 1993; updated using European Journal of Political Research, various years, and election data available at: http://www-public.rz.uni-duesseldorf.de/~nordsiew/indexe.html.//

Number of Government Parties (GOVPTYS)

Number of parties in government (cabinet). Constructed in G. Bingham Powell, , Jr. and Guy D. Whitten, “A Cross-National Analysis of Economic Voting,” American Journal of Political Science 37(2): 391-414, 1993;  updated using European Journal of Political Research, Political Data Yearbook, various years.  Updated through 1997.

Endogenous Election (ENDELECT)

dummy if the parliamentary election at time t was endogenous.

Exogenous Election (EXELECTN)

dummy if the parliamentary election at time t was exogenous.

Legislative Center of Gravity (LEGISCG)

see CABCG above.  Same scale applied to entire parliament at time t, based on number of seats won by each party in the most recent election.  In election years, data were calculated as the weighted average of the pre-election parliament and the post-election parliament.  Data calculated from European Journal of Political Research Political Data Yearbook, various years, and election/seat data available at: http://www-public.rz.uni-duesseldorf.de/~nordsiew/indexe.html.  Updated through 1997.

Government Ideology (GOVIDEO)

Constructed by Powell and Whitten (AJPS 37: 391-414) from average ratings of parties on a left-right scale in Castles and Mair's survey of country experts (EJPR 12: 73-88).  For coalition governments, each party's score was weighted by the % of legislative seats it held in the cabinet.  Data calculated from European Journal of Political Research Political Data Yearbook, various years.  For years with multiple governments, data were calculated as a weighted average of the “old” and “new” governments. Updated through 1997.

Government Ideology Gap (GOVGAP)

Difference between the government parties with the lowest and highest Castles and Mairs ratings.  Again, for years with multiple governments, data calculated as weighted average of old and new governments.  Updated through 1997.

VOTET, VOTETM1, VOTETM2

Percentage of legislative seats won by the governments parties in the elections at times t, t-1, and t-2 (the three most recent elections).  For election years, data are calculated as the weighted average of the outgoing and incoming governments' scores for all three variables.  Updated through 1997.

BICAMOPP

dummy =1 when governing party did NOT control majority in second legislative chamber.  Constructed from Powell and Whitten.

OPPCHAIR

dummy=1 when legislature had strong committee system with proportionate sharing.  Constructed from Powell and Whitten.

WEAKPTY

dummy=1 if major parties lack internal cohesion in voting behavior.  Constructed from Powell and Whitten.

3.2.5 Political Economy Variables

Union Density (UNDENS)

union density, adjusted for unemployment and self-employment.  Data is Visser, from ADD G&W LINK.  This data is available through 1992-3, and replaces the old data, which was net union density (adjusted only for self-employment, not unemployment) and was available only through 1990.

Member of Snake/ERM (SNAKEMS)

Dichotomous variable = 1 if country is a member of either Snake or ERM, 0 if not. Data obtained from the BIS Annual Reports, various years, and updated using data available at http://www.ft.com/euro

Central Bank Independence (CBI)

An index of central bank independence, running from 0 (least independent to 1 (most independent). Data from Cukierman, Webb and Neyapti (1992).

3.2.6 VARIABLES NOT YET UPDATED

Capital Controls (CLOSED)

A measure of capital controls constructed by Dennis Quinn, described in  Dennis Quinn, “The Correlates of Change in International Financial Regulation” in American Political Science Review 91(3): 531-552, 1997. His 15 point-scale measures “openness;” it is inverted here so that a higher number means more capital controls. Data obtained from the author. NOTE: not yet updated.

Manufactured Exports to DM Zone as % of GDP

Value of manufactured (SITC codes 6-8) exports to Germany, Belgium, Luxembourg and the Netherlands as a percentage of  GDP. Data supplied by the UN from various years of their Yearbook of International Trade Statistics.  NOTE: not yet updated.

Government Duration (GOVDUR)

number of months government was in office.  NOTE: current data are incorrect – need to update.

Manufactured exports to the DM Zone (DMMFG1)

Value of manufactured (SITC codes 6-8) exports to Germany and Benelux as % of GDP.

Manufactured exports to the EC/EU (ECMFG1)

Value of manufactured (SITC codes 6-8) exports to the EC/EU as % of GDP.

Public sector employment (PUBEMP)

emailing Garrett for info on updated data availability.

Public sector unionization (PUBUNION)

emailing Garrett for info on updated data availability.

Foreign direct investment data

ECFDI and DMFDI