## _wp16230

## Source details

**Canonical URL:** [_wp16230](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16230.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16230.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16230.pdf.json)

---

### Key findings and summary
- General empirical result: policy makers that take decisions under fixed exchange rate regimes are associated with less fiscal discipline.
- Political context matters: strong political environments (longer political horizon and higher political cohesion) help improve fiscal discipline, especially under flexible exchange rate regimes.
- Interaction effect: policy makers who don’t face immediate elections and who operate without political fragmentation are associated with better fiscal performance, particularly when operating within a flexible exchange rate regime.

### Conceptual contributions (political dimensions)
- Political Horizon
  - Defined as the time the policy-makers have before forthcoming elections (or regime change in autocracies).
  - Politicians facing longer horizons have less incentive to overspend and are associated with more fiscal discipline.
- Political Cohesion
  - Captures the number of political actors participating in budgetary decisions (parties in government or opposition, interest groups, veto players).
  - More cohesive political environments reduce conflicting spending demands and are associated with tighter fiscal discipline.

### Theoretical framework (model structure and mechanisms)
- Model setting: two periods and three economic agents — a monetary authority (decides the exchange rate) and two political players (authority over public spending).
- Central bank objective: minimize deviations of the nominal exchange rate E from target E* and output Y from target Y*; loss function incorporates discount factor ߚெ and weight ܽெ with ܽெ > 1.
- Political players objective: maximize constituency output; loss function for constituency 1: L_p = - ln(Y1,1) - ߚি * ln(Y1,2), where ߚি is the policymakers’ discount factor and proxies political horizon (smaller ߚি implies shorter horizon).
- Fiscal game timing:
  - Period 0: monetary authority sets a fixed exchange rate ܧଵ (not endogenized).
  - Period 1: two political players choose transfers to their constituencies (T1 and T2), which increase output in period 1 but generate debt D = T1 + T2 to be repaid in period 2.
  - Period 2: debt is fully repaid; burden shared by constituencies with weights ߙଵ and ߙଶ = 1 − ߙଵ.
- Output equations:
  - Period 1 outputs: Y1,1 = f(E1) + T1; Y2,1 = f(E1) + T2
  - Period 2 outputs: Y1,2 = f(E2) + ߙଵ·D; Y2,2 = f(E2) + ߙଶ·D
- Mechanism: under fixed exchange rates, short political horizons and political fragmentation can induce fiscal profligacy because immediate electoral or coalition incentives dominate future repayment costs; under flexible rates, immediate economic punishment (currency depreciation, inflation) raises the cost of fiscal laxity, particularly when political horizons are longer and cohesion is higher.

### Literature synthesis
- Traditional view: fixed exchange rates encourage fiscal discipline via anticipation of high costs of abandoning the peg.
- Dynamic counterargument (Tornell and Velasco; Sun): fixed pegs can induce fiscal indiscipline when policymakers heavily discount the future because economic costs of indiscipline are delayed; flexible rates deliver more immediate punishment.
- Paper’s contribution: integrates political economy dimensions (electoral timing and political cohesion) into the analysis to explain mixed empirical results on the exchange rate–fiscal discipline relationship.

### Empirical approach and scope
- Sample: unbalanced panel of 79 countries (31 advanced, 48 emerging and low income) between 1975 and 2012.
- Baseline regression (equation (9)): dependent variable itFD is primary balance (percent of GDP). Key regressors:
  - itER: dummy = 1 if fixed exchange rate regime, 0 if flexible (Reinhart and Rogoff, 2004 classification, values 1–8 → fixed).
  - POL: political factors constructed by PCA into two components: Horizon and Cohesion.
  - Controls: real GDP growth, trade openness (exports+imports share of GDP), terms of trade, log of total reserves (USD), private credit (% GDP), CPI inflation.
- Political composite variables (PCA):
  - Political Horizon factor: years in office; years left in current term; party of chief executive with long tradition in office; months to next election. First principal component retained; explains almost 40 percent of variance in standardized data.
  - Political Cohesion factor: margin of majority; cabinet strength; executive control of all houses; weak opposition. First principal component retained; Cohesion factor explains more variance (Table 2 shows % explained 0.39 for Horizon, 0.69 for Cohesion).

### Theoretical propositions (debt, central bank accommodation, and political parameters)
- Two political economy parameters determine indebtedness:
  - Political Horizon: the discount factor ߚி; a longer Political Horizon (lower discount of the future) leads policymakers to value future utility more and reduces current transfers and indebtedness.
  - Political Cohesion: the share of debt each fiscal actor repays (more even sharing, e.g., each pays 50 percent, implies higher cohesion), which reduces excess indebtedness.
- Central bank reaction: higher debt in period 1 → higher exchange rate depreciation in period 2. Accommodation is less than one-for-one because ܽெ > 1 and (ߙଵߛ∙ଵ + ߙଶߛ∙ଶ) generally < 1. As ܽெ → infinity, ܧଶ → E* and there is no debt accommodation (fully credible fixed exchange rate).
- Fiscal reaction: a 1 monetary unit increase in transfers by one player reduces the other's transfers by less than one unit → existence of equilibrium.
- Propositions:
  - Proposition 1: An decrease in the discount factor, ߚி, (i.e., longer political Horizon) leads to lower indebtedness.
  - Proposition 2: An increase in political Cohesion (more even distribution of debt burden) leads to lower indebtedness.

### Baseline empirical results (fixed effects, selected estimates)
- Fixed ER coefficient: negative in all specifications; example: Fixed ER = -0.432** (standard error 0.180) in Table 3, specification (1).
- Controls (examples and exact values reported in source):
  - Growth: 0.087*** (0.019)
  - Trade Openness: 0.017*** (0.005)
  - Terms of Trade: 0.017*** (0.003)
  - Reserves: 0.438*** (0.093)
  - Inflation (where included): 0.034*** (0.011)
- Political variables (examples):
  - Horizon: 0.346*** (0.096) in Table 3.
  - Cohesion: 0.241* (0.142) in Table 3, column (6).

### Interaction effects (Table 4 highlights)
- Interaction terms between Fixed ER and political variables are negative:
  - Horizon*Fixed ER: -0.435*** (0.167) in Table 4, specification (1).
  - Cohesion*Fixed ER: -0.397* (0.231) in Table 4, specification (2).
- Interpretation:
  - Flexible exchange rates and stronger politics (longer Horizon, higher Cohesion) are associated with better fiscal positions.
  - Exchange rate regime is quantitatively important; politics attenuate but do not fully reverse the adverse effect of fixed rates.
  - Positive effects of Horizon and Cohesion on fiscal performance are amplified under flexible ERs.

### Robustness checks and endogeneity (selected results)
- Alternative estimators (Table 5): FGLS, Driscoll-Kraay, and outlier-robust MM-estimator yield similar findings; Fixed ER ranges from -0.579*** to -0.956*** across estimators and specs.
- Horizon remained positive and significant in many robustness specs (examples: 0.215***, 0.315***, 0.344*** in Table 5 columns).
- Horizon*Fixed ER often negative and significant (examples: -0.247** (0.111) and -0.423*** (0.157) in Table 5).
- Instrumental variables (Table 6): IV estimates confirm baseline:
  - Fixed ER negative and significant (e.g., -0.730*** (0.203) to -0.888*** (0.227)).
  - Horizon positive and significant in IV (e.g., 0.365*** (0.111); 0.536*** (0.136)).
  - Horizon*Fixed ER negative and significant in IV (e.g., -0.431** (0.200)).
  - Instruments include constraints on the executive and political constraints (polcon).

### Alternative dependent variable: fiscal consolidations (Table 7)
- Dependent variable: binary indicator of fiscal consolidation episodes following Alesina and Ardagna (1998) criteria; logistic fixed-effects estimation.
- Fixed ER decreases likelihood of fiscal consolidation:
  - All Fiscal Consolidations: Fixed ER = -0.405*** (0.149) in column (1).
  - Successful Fiscal Consolidations: Fixed ER = -1.173** (0.556), -1.720*** (0.646), -3.353* (1.877) across different columns.
- Horizon generally not significant for consolidation incidence; Cohesion positive and significant in some specifications for consolidations (example: 0.782*** (0.189) in column (6)); Cohesion*Fixed ER not robustly significant.

### Alternative ER definition and sensitivity (Table 8 and notes)
- Alternative fixed ER dummy (Fixed ER-new): defined as 5-year rolling volatility of the real effective exchange rate < 1/3 of 5-year rolling average volatility.
  - Fixed ER-new remains associated with lower fiscal discipline; example: Fixed ER-new = -0.952*** (0.356) in Table 8, specification (3) when Cohesion controlled for.
- Results robust to alternative definitions of dependent variable (nominal levels, real per capita) and to different consolidation identification methods (Giavazzi and Pagano (1996) used as robustness check).

### Key empirical conclusions
- Being in or moving to a fixed exchange rate regime is associated with less fiscal discipline (negative and often significant coefficient on Fixed ER across specifications and estimators).
- Political factors matter: longer political Horizon and higher political Cohesion are associated with better fiscal discipline (positive coefficients).
- Interaction: the beneficial effects of political Horizon and Cohesion on fiscal discipline are stronger under flexible exchange rate regimes; political strength attenuates but does not reverse the adverse fiscal effects of fixed exchange rates.
- Results are robust to alternative estimators, outlier treatments, instrumental variables addressing endogeneity, alternative ER definitions, and alternative dependent variable definitions (fiscal consolidations).

### Synthetic Control Methodology (SCM) — design and implementation
- Outcome variable: primary budget balance (in percent of GDP).
- Predictor variables used to match primary balance behavior (3-year moving averages):
  - The primary balance in 1981, 1993 and 2000
  - Total reserves minus gold (US Dollars Millions)
  - Real GDP Growth (%)
  - Terms of Trade, US Dollars (Index)
  - Trade Openness (% in U.S. dollars)
  - Change in Nominal Exchange Rate
  - Change in Debt-to GDP ratio
- Year of the event: 2001 (last year before actual circulation of the single currency).
- Donor pool: Denmark, Greece, Iceland, Israel, Norway, Sweden, Switzerland, Turkey and the UK.
- Country cases (Euro-11): Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Portugal and Spain.
- SCM estimation algorithm (nested optimization; iterative):
  - Step 1: choose initial vector of variable weights V; choose country weight vector w* to minimize ‖X1 − X0 w‖ subject to 0 ≤ weights ≤ 1.
  - Step 2: given w*, choose V* to minimize MSPE of the outcome over pre-event periods: V* = argmin_V Σ_{t ∈ pre} (Y1_t − Y0_t w*)' (Y1_t − Y0_t w*).
  - Iterate (1) and (2) until convergence; use resulting weights to construct synthetic comparator and create counterfactual post-event path.
  - Step 3: compare actual post-event outcome series with synthetic comparator; difference is the estimated impact.

### SCM empirical findings (selected results and interpretations)
- Euro-11 as a group, after moving to fixed exchange rates, maintained a lower primary balance than its synthetic counterfactual.
- Political environment moderates fiscal outcomes after moving to fixed exchange rates:
  - Countries with high relative scores in at least one political dimension (Horizon or Cohesion), e.g., Austria (higher-than-average Horizon) or Portugal (higher-than-average Cohesion), experienced less fiscal slippage and display a narrower distance to their synthetic counterfactuals.
  - Countries with lower-than-average scores in both political dimensions, e.g., Belgium, saw their distance to the counterfactual increase.
  - Countries with higher-than-average scores in both political Horizon and Cohesion, e.g., Spain, managed to improve fiscal discipline well above their synthetic counterfactual; Spain robustness check using structural balance yields similar conclusion.
- Policy interpretation: stronger political environments can partially offset the negative fiscal effect of moving to fixed exchange rates; weak political institutions cannot be fully substituted by hard pegs.

### Robustness evidence (select econometric results from Appendix)
- Table A1 selected coefficient estimates (exact values as reported):
  - Fixed ER: -0.926*** (spec (1)), -0.385 (spec (2)), -0.424 (spec (3)), -0.618* (spec (4)), -0.233 (spec (5)), -0.881*** (spec (6)), -0.369 (spec (7)), -0.411 (spec (8)), -0.603* (spec (9)), -0.206 (spec (10)).
  - Growth: -0.006 (spec (1)), 0.030 (spec (2)), 0.030 (spec (3)), 0.042 (spec (4)), 0.041 (spec (5)), -0.005 (spec (6)), 0.030 (spec (7)), 0.030 (spec (8)), 0.043 (spec (9)), 0.042 (spec (10)).
  - Trade Openness: -0.034*** (spec (1)), -0.022*** (spec (2)), -0.022*** (spec (3)), -0.039*** (spec (4)), -0.038*** (spec (5)), -0.034*** (spec (6)), -0.022*** (spec (7)), -0.022*** (spec (8)), -0.039*** (spec (9)), -0.038*** (spec (10)).
  - Terms Of Trade: -0.006 (spec (1)), -0.002 (spec (2)), -0.002 (spec (3)), -0.008 (spec (4)), -0.008 (spec (5)), -0.006 (spec (6)), -0.002 (spec (7)), -0.002 (spec (8)), -0.008 (spec (9)), -0.008 (spec (10)).
  - Reserves: 0.625*** (spec (1)), 0.460*** (spec (2)), 0.462*** (spec (3)), 0.832*** (spec (4)), 0.840*** (spec (5)), 0.592*** (spec (6)), 0.427*** (spec (7)), 0.429*** (spec (8)), 0.813*** (spec (9)), 0.821*** (spec (10)).
  - Horizon: -0.154 (std. err. 0.104) in one specification; -0.121 (std. err. 0.140) in another.
  - Horizon*Fixed ER: -0.088 (std. err. 0.202) in one reported interaction; -0.093 (std. err. 0.201) in another.
  - Cohesion: 0.909*** (std. err. 0.291), 0.746** (std. err. 0.307) in two specifications; 0.943*** (std. err. 0.288), 0.775** (std. err. 0.304) in parallel specs.
  - Cohesion*Fixed ER: 0.566 (std. err. 0.387) in one interaction; 0.584 (std. err. 0.383) in another.
- Observations and fit reported:
  - Observations: 1,121 (specs (1) and (6)), 896 (specs (2), (3), (7), (8)), 800 (specs (4), (5), (9), (10)).
  - R-squared: 0.740 (spec (1)), 0.762 (specs (2) and (3)), 0.737 (spec (4)), 0.738 (spec (5)), 0.741 (spec (6)), 0.756 (specs (7) and (8)), 0.736 (spec (9)), 0.737 (spec (10)).
- Notes: Estimation with country and time effects (omitted for parsimony). Robust standard errors clustered at the country level in parenthesis. Constant term omitted. *, **, *** denote statistical significance at the 1, 5 and 10 percent levels, respectively.

### Main conclusions and policy recommendations
- Core empirical conclusion: Contrary to the traditional argument, flexible (not fixed) exchange rate regimes are associated with more fiscal discipline.
- Political context matters: the fiscal implications of exchange rate regime choice operate within a specific political environment; strong political environments with long Horizon and high Cohesion are associated with better fiscal performance.
- Policy lessons:
  - If policymakers operate in weak political contexts, flexible exchange rates are the preferred option as they are best suited to secure enhanced fiscal discipline.
  - The positive effect of flexible exchange rates on fiscal discipline can be maximized when political environments are strong (long political horizon and high political cohesion).
  - In mixed political contexts, policymakers face a difficult choice: moving to a fixed exchange rate regime may negatively impact fiscal performance in a way that political institutions are unable to attenuate.

*Source: _wp16230 - 2.  How  much  each  fiscal  policymaker  engages  in  transfers  today  depends  on  the  amount  of  future (PDF chapter/section) and _wp16230 - introduction of the single currency, using data on a set of variables from the years prior to the event.*

### References ________________________________________________________________________________________ 29

### _wp16230 - References ________________________________________________________________________________________ 29

### Key findings and summary
- General empirical result: policy makers that take decisions under fixed exchange rate regimes are associated with less fiscal discipline.
- Political context matters: strong political environments (longer political horizon and higher political cohesion) help improve fiscal discipline, especially under flexible exchange rate regimes.
- Interaction effect: policy makers who don’t face immediate elections and who operate without political fragmentation are associated with better fiscal performance, particularly when operating within a flexible exchange rate regime.

### Conceptual contributions (political dimensions)
- Political Horizon
  - Defined as the time the policy-makers have before forthcoming elections (or regime change in autocracies).
  - Politicians facing longer horizons have less incentive to overspend and are associated with more fiscal discipline.
- Political Cohesion
  - Captures the number of political actors participating in budgetary decisions (parties in government or opposition, interest groups, veto players).
  - More cohesive political environments reduce conflicting spending demands and are associated with tighter fiscal discipline.

### Theoretical framework (model structure and mechanisms)
- Model setting: two periods and three economic agents — a monetary authority (decides the exchange rate) and two political players (authority over public spending).
- Central bank objective: minimize deviations of the nominal exchange rate E from target E* and output Y from target Y*; loss function incorporates discount factor ߚெ and weight ܽெ with ܽெ > 1.
- Political players objective: maximize constituency output; loss function for constituency 1: L_p = - ln(Y1,1) - ߚি * ln(Y1,2), where ߚি is the policymakers’ discount factor and proxies political horizon (smaller ߚি implies shorter horizon).
- Fiscal game timing:
  - Period 0: monetary authority sets a fixed exchange rate ܧଵ (not endogenized).
  - Period 1: two political players choose transfers to their constituencies (T1 and T2), which increase output in period 1 but generate debt D = T1 + T2 to be repaid in period 2.
  - Period 2: debt is fully repaid; burden shared by constituencies with weights ߙଵ and ߙଶ = 1 − ߙଵ.
- Output equations:
  - Period 1 outputs: Y1,1 = f(E1) + T1; Y2,1 = f(E1) + T2
  - Period 2 outputs: Y1,2 = f(E2) + ߙଵ·D; Y2,2 = f(E2) + ߙଶ·D
- Mechanism: under fixed exchange rates, short political horizons and political fragmentation can induce fiscal profligacy because immediate electoral or coalition incentives dominate future repayment costs; under flexible rates, immediate economic punishment (currency depreciation, inflation) raises the cost of fiscal laxity, particularly when political horizons are longer and cohesion is higher.

### Literature synthesis (how paper positions itself)
- Traditional view: fixed exchange rates encourage fiscal discipline via anticipation of high costs of abandoning the peg.
- Dynamic counterargument (Tornell and Velasco; Sun): fixed pegs can induce fiscal indiscipline when policymakers heavily discount the future because economic costs of indiscipline are delayed; flexible rates deliver more immediate punishment.
- Paper’s contribution: integrates political economy dimensions (electoral timing and political cohesion) into the analysis to explain mixed empirical results on the exchange rate–fiscal discipline relationship.

### Empirical approach and scope (as described)
- Uses the broadest possible sample yet available (sample scope statement in introduction).
- Empirical strategy: incorporate political composite variables (political horizon and political cohesion) into estimation of how exchange rate regimes interact with fiscal discipline.
- Additional empirical elements mentioned: fixed effects regressions, interaction terms, alternative estimators, instrumental variables for endogeneity, logit estimations for discretionary fiscal consolidations, robustness checks, and synthetic control analysis for the euro area (as listed in figures, tables, and appendices).

### Policy implications
- Exchange rate regime choice interacts with domestic political conditions; regime choice alone is an incomplete policy lever for fiscal discipline.
- In contexts with weak political institutions (short horizons, high fragmentation), flexible exchange rates can act as a partial substitute for weaker accountability mechanisms by exposing policymakers to more immediate economic costs from fiscal profligacy.
- Strengthening political cohesion and lengthening political horizons (institutional reforms that enhance policymaker accountability and reduce short-term electoral pressures) complement flexible exchange rate regimes in promoting fiscal discipline.
- Hard and credible pegs are not a guaranteed substitute for strong national institutions ensuring accountability and transparency (Eurozone experience noted).

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16230.pdf*

### 2.  How  much  each  fiscal  policymaker  engages  in  transfers  today  depends  on  the  amount  of  future

### _wp16230 - 2.  How  much  each  fiscal  policymaker  engages  in  transfers  today  depends  on  the  amount  of  future

### Theoretical framework: debt, central bank accommodation, and political parameters
- Central bank in period 2 cares for output and the exchange rate and will default on the exchange rate; the extent of accommodation of fiscal profligacy depends on ܽ
ெ
.
- Two political economy parameters determine indebtedness:
  - Political Horizon: the discount factor ߚ
ி
; a longer political Horizon (lower discount of the future) leads policymakers to value future utility more and reduces current transfers and indebtedness.
  - Political Cohesion: the share of debt each fiscal actor repays (more even sharing, e.g., each pays 50 percent, implies higher cohesion), which reduces excess indebtedness.
- Central bank reaction function (period 2 devaluation as function of debt D) implies:
  - Higher debt in period 1 → higher exchange rate depreciation in period 2.
  - Accommodation is less than one-for-one because ܽ
ெ
 > 1 and (ߙ
ଵ
ߛ∙
ଵ
൅
ߙ
ଶ
ߛ∙
ଶ
) generally < 1.
  - As ܽ
ெ
 → infinity, ܧ
ଶ
 → E* and there is no debt accommodation (fully credible fixed exchange rate).

### Fiscal reaction functions and equilibrium (period 1 transfers)
- Policymakers choose transfers anticipating central bank behavior; each constituency’s transfers depend on the other's transfers (equations (7) and (8)).
- A 1 monetary unit increase in transfers by one player reduces the other's transfers by less than one unit → existence of equilibrium.
- Two political-economy propositions derived:
  - Proposition 1: An decrease in the discount factor, ߚ
ி
, (i.e., longer political Horizon) leads to lower indebtedness.
  - Proposition 2: An increase in political Cohesion (more even distribution of debt burden) leads to lower indebtedness.

### Empirical analysis: data, variables, and specification
- Data: unbalanced panel of 79 countries (31 advanced, 48 emerging and low income) between 1975 and 2012.
- Baseline regression (equation (9)): dependent variable itFD is primary balance (percent of GDP). Key regressors:
  - itER: dummy = 1 if fixed exchange rate regime, 0 if flexible (Reinhart and Rogoff, 2004 classification, values 1–8 → fixed).
  - POL: political factors constructed by PCA into two components: Horizon and Cohesion.
  - Controls: real GDP growth, trade openness (exports+imports share of GDP), terms of trade, log of total reserves (USD), private credit (% GDP), CPI inflation.
- Political composite variables (PCA):
  - Political Horizon factor: years in office; years left in current term; party of chief executive with long tradition in office; months to next election. First principal component retained; explains almost 40% of variance in standardized data.
  - Political Cohesion factor: margin of majority; cabinet strength; executive control of all houses; weak opposition. First principal component retained; Cohesion factor explains more variance (Table 2 shows % explained 0.39 for Horizon, 0.69 for Cohesion).
- Stylized facts:
  - Long Horizon: average primary surplus was 0.01 percent of GDP under fixed ER and 0.3 percent of GDP under flexible ER.
  - High Cohesion: average primary deficit was -0.2 percent of GDP under fixed ER and -0.1 percent of GDP under flexible ER.

### Baseline empirical results (fixed effects, Table 3 highlights)
- Fixed ER coefficient: negative in all specifications; e.g., Fixed ER = -0.432** (specification (1), standard error (0.180)).
- Controls typically with expected signs: Growth positive and significant (e.g., 0.087*** (0.019)); Trade Openness positive and significant (e.g., 0.017*** (0.005)); Terms of Trade positive and significant (e.g., 0.017*** (0.003)); Reserves positive and significant in many specs (e.g., 0.438*** (0.093)); Inflation positive where included (e.g., 0.034*** (0.011)).
- Political variables:
  - Horizon: positive and significant (e.g., 0.346*** (0.096) in Table 3).
  - Cohesion: positive and in some specs significant (e.g., 0.241* (0.142) in Table 3, column (6)).

### Interaction effects and interpretation (Table 4 highlights)
- Interaction terms between Fixed ER and political variables are negative:
  - Horizon*Fixed ER: -0.435*** (0.167) in Table 4, specification (1).
  - Cohesion*Fixed ER: negative (e.g., -0.397* (0.231) in Table 4, specification (2)).
- Interpretation:
  - Flexible exchange rates and stronger politics (longer Horizon, higher Cohesion) are associated with better fiscal positions.
  - Exchange rate regime is quantitatively important; politics attenuate but do not fully reverse the adverse effect of fixed rates.
  - Positive effects of Horizon and Cohesion on fiscal performance are amplified under flexible ERs.

### Robustness checks and endogeneity (Tables 5–6 highlights)
- Alternative estimators:
  - FGLS, Driscoll-Kraay, and outlier-robust MM-estimator yield similar findings: fixed ER associated with less fiscal discipline (e.g., Table 5 Fixed ER ranges from -0.579*** to -0.956*** across estimators and specs).
  - Horizon remains positive and significant in many robustness specs (e.g., 0.215***, 0.315***, 0.344*** in Table 5 columns).
  - Horizon*Fixed ER often negative and significant (e.g., -0.247** (0.111) and -0.423*** (0.157) in Table 5).
- Instrumental variables (Table 6):
  - IV estimates confirm baseline: Fixed ER negative and significant (e.g., -0.730*** (0.203) to -0.888*** (0.227)).
  - Horizon positive and significant in IV (e.g., 0.365*** (0.111); 0.536*** (0.136) in different IV specs).
  - Horizon*Fixed ER negative and significant in IV (e.g., -0.431** (0.200) in Table 6).
  - Instruments include constraints on the executive and political constraints (polcon).

### Alternative dependent variable: fiscal consolidations (Table 7 highlights)
- Dependent variable: binary indicator of fiscal consolidation episodes following Alesina and Ardagna (1998) criteria; logistic fixed-effects estimation.
- Fixed ER decreases likelihood of fiscal consolidation:
  - All Fiscal Consolidations: Fixed ER coefficients negative and significant (e.g., -0.405*** (0.149) in column (1)).
  - Successful Fiscal Consolidations: larger negative magnitudes (e.g., -1.173** (0.556), -1.720*** (0.646), -3.353* (1.877) across columns).
- Horizon generally not significant for consolidation incidence; Cohesion positive and significant in some specifications for consolidations (e.g., 0.782*** (0.189) in column (6)) but Cohesion*Fixed ER not robustly significant.

### Robustness to alternative ER definition and additional sensitivity (Table 8 and notes)
- Alternative fixed ER dummy (Fixed ER-new): defined as 5-year rolling volatility of the real effective exchange rate < 1/3 of 5-year rolling average volatility.
  - Fixed ER-new remains associated with lower fiscal discipline; e.g., Fixed ER-new = -0.952*** (0.356) in Table 8, specification (3) when Cohesion controlled for.
- Results robust to alternative definitions of dependent variable (nominal levels, real per capita) and to different consolidation identification methods (Giavazzi and Pagano (1996) used as robustness check).

### Key empirical conclusions
- Being in or moving to a fixed exchange rate regime is associated with less fiscal discipline (negative and often significant coefficient on Fixed ER across specifications and estimators).
- Political factors matter: longer political Horizon and higher political Cohesion are associated with better fiscal discipline (positive coefficients).
- Interaction: the beneficial effects of political Horizon and Cohesion on fiscal discipline are stronger under flexible exchange rate regimes; political strength attenuates but does not reverse the adverse fiscal effects of fixed exchange rates.
- Results are robust to alternative estimators, outlier treatments, instrumental variables addressing endogeneity, alternative ER definitions, and alternative dependent variable definitions (fiscal consolidations).

_Italic line: Source: _wp16230 - 2.  How  much  each  fiscal  policymaker  engages  in  transfers  today  depends  on  the  amount  of  future (PDF chapter/section)._

### introduction of the single currency, using data on a set of variables from the years prior to the event,

### _wp16230 - introduction of the single currency, using data on a set of variables from the years prior to the event,

### Synthetic Control Methodology (SCM) — design and implementation
- Counterfactual (synthetic unit) constructed as a weighted average of the primary budget balance in countries with similar characteristics but unaffected by the event.
- Country weights are chosen to minimize the distance between the country under consideration and its counterfactual in terms of the primary balance variable and its predictors (Abadie and others, 2010).
- Outcome variable: primary budget balance (in percent of GDP).
- Predictor variables used to match primary balance behavior:
  - The primary balance in 1981, 1993 and 2000
  - Total reserves minus gold (US Dollars Millions)
  - Real GDP Growth (%)
  - Terms of Trade, US Dollars (Index)
  - Trade Openness (% in U.S. dollars)
  - Change in Nominal Exchange Rate
  - Change in Debt-to GDP ratio
- All variables take 3-year moving average values to smooth fluctuations.
- Year of the event: 2001 (last year before actual circulation of the single currency).
- Donor pool (comparator countries): Denmark, Greece, Iceland, Israel, Norway, Sweden, Switzerland, Turkey and the UK (OECD European countries that did not join the Eurozone in 1999; transition economies excluded).
- Country cases (Euro-11): Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Portugal and Spain.

- SCM estimation algorithm (nested optimization; iterative):
  - Step 1: Choose an initial vector of weights of the predictor variables V; choose country weight vector w* to minimize distance ‖X1 − X0 w‖ where X1 and X0 are matrices of predictor variables for the unit of interest and comparator units, subject to 0 ≤ weights ≤ 1. (Equation (1) in text.)
  - Step 2: Given w*, choose variable weight matrix V* (positive definite and diagonal) to minimize the mean square prediction error (MSPE) of the outcome variable over pre-event periods: V* = argmin_V Σ_{t ∈ pre} (Y1_t − Y0_t w*)' (Y1_t − Y0_t w*). (Equation (2) in text.)
  - Iterate (1) and (2) until w* and V* converge.
  - Use resulting weights to construct synthetic comparator and create counterfactual post-event path.
  - Step 3: Compare actual post-event outcome series with synthetic comparator; difference is the estimated impact of the event (assuming other factors controlled).

### SCM empirical findings (selected results and interpretations)
- Euro-11 as a group, in the aftermath of moving to fixed exchange rates, maintained a lower primary balance than its synthetic counterfactual.
- Political environment moderates fiscal outcomes after moving to fixed exchange rates:
  - Countries with high relative scores in at least one political dimension (Horizon or Cohesion), e.g., Austria (higher-than-average Horizon) or Portugal (higher-than-average Cohesion), experienced less fiscal slippage and display a narrower distance to their synthetic counterfactuals.
  - Countries with lower-than-average scores in both political dimensions, e.g., Belgium, saw their distance to the counterfactual increase (i.e., worse fiscal discipline relative to synthetic control).
  - Countries with higher-than-average scores in both political Horizon and Cohesion, e.g., Spain, managed to improve fiscal discipline well above their synthetic counterfactual, indicating that strong politics partially compensates for the negative fiscal effect of moving to fixed exchange rates.
- Spain robustness check: analysis replicated using structural balance to control for the real estate bubble’s effect on fiscal revenues during the late nineties; results remain unchanged — Spain still performed better after the introduction of the euro, though it "could have adjusted further."

### Main conclusions and policy recommendations
- Core empirical conclusion: Contrary to the traditional argument, flexible (not fixed) exchange rate regimes are associated with more fiscal discipline.
- Political context matters: the fiscal implications of exchange rate regime choice operate within a specific political environment; strong political environments with long Horizon and high Cohesion are associated with better fiscal performance.
- Policy lessons:
  - If policymakers operate in weak political contexts, flexible exchange rates are the preferred option as they are best suited to secure enhanced fiscal discipline.
  - The positive effect of flexible exchange rates on fiscal discipline can be maximized when political environments are strong (long political horizon and high political cohesion).
  - In mixed political contexts, policymakers face a difficult choice: moving to a fixed exchange rate regime may negatively impact fiscal performance in a way that political institutions are unable to attenuate.

### Robustness evidence (select econometric results)
- Table A1: Robustness to alternative measures of fiscal discipline; dependent variables include Nominal Primary Balance, Real Primary Balance, and Primary Balance per capita across multiple specifications (columns (1) to (10)).
- Selected coefficient estimates (exact values reported):
  - Fixed ER: -0.926*** (spec (1)), -0.385 (spec (2)), -0.424 (spec (3)), -0.618* (spec (4)), -0.233 (spec (5)), -0.881*** (spec (6)), -0.369 (spec (7)), -0.411 (spec (8)), -0.603* (spec (9)), -0.206 (spec (10)).
  - Growth: -0.006 (spec (1)), 0.030 (spec (2)), 0.030 (spec (3)), 0.042 (spec (4)), 0.041 (spec (5)), -0.005 (spec (6)), 0.030 (spec (7)), 0.030 (spec (8)), 0.043 (spec (9)), 0.042 (spec (10)).
  - Trade Openness: -0.034*** (spec (1)), -0.022*** (spec (2)), -0.022*** (spec (3)), -0.039*** (spec (4)), -0.038*** (spec (5)), -0.034*** (spec (6)), -0.022*** (spec (7)), -0.022*** (spec (8)), -0.039*** (spec (9)), -0.038*** (spec (10)).
  - Terms Of Trade: -0.006 (spec (1)), -0.002 (spec (2)), -0.002 (spec (3)), -0.008 (spec (4)), -0.008 (spec (5)), -0.006 (spec (6)), -0.002 (spec (7)), -0.002 (spec (8)), -0.008 (spec (9)), -0.008 (spec (10)).
  - Reserves: 0.625*** (spec (1)), 0.460*** (spec (2)), 0.462*** (spec (3)), 0.832*** (spec (4)), 0.840*** (spec (5)), 0.592*** (spec (6)), 0.427*** (spec (7)), 0.429*** (spec (8)), 0.813*** (spec (9)), 0.821*** (spec (10)).
  - Horizon: -0.154 (std. err. 0.104) in one specification; -0.121 (std. err. 0.140) in another.
  - Horizon*Fixed ER: -0.088 (std. err. 0.202) in one reported interaction; -0.093 (std. err. 0.201) in another.
  - Cohesion: 0.909*** (std. err. 0.291), 0.746** (std. err. 0.307) in two specifications; 0.943*** (std. err. 0.288), 0.775** (std. err. 0.304) in parallel specs.
  - Cohesion*Fixed ER: 0.566 (std. err. 0.387) in one interaction; 0.584 (std. err. 0.383) in another.
- Observations and fit:
  - Observations: 1,121 (specs (1) and (6)), 896 (specs (2), (3), (7), (8)), 800 (specs (4), (5), (9), (10)).
  - R-squared: 0.740 (spec (1)), 0.762 (specs (2) and (3)), 0.737 (spec (4)), 0.738 (spec (5)), 0.741 (spec (6)), 0.756 (specs (7) and (8)), 0.736 (spec (9)), 0.737 (spec (10)).
- Note: Estimation with country and time effects (omitted for parsimony). Robust standard errors clustered at the country level in parenthesis. Constant term omitted. *, **, *** denote statistical significance at the 1, 5 and 10 percent levels, respectively.

### Appendix 2 — SCM procedural steps (summary)
- Step 1: Choose country cases (Euro-11), donor pool (OECD European non-Euro-11 excluding transition economies), and explanatory variables as listed above.
- Step 2: Compute synthetic series by iteratively optimizing country weights and predictor-variable weights to minimize pre-event MSPE, subject to weight constraints (0–1).
- Step 3: Use synthetic comparator to construct counterfactual post-event outcome path and estimate event impact by difference from actual series.

*Source: _wp16230 - introduction of the single currency, using data on a set of variables from the years prior to the event,*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16230.pdf_