## _wp16171

## Source details

**Canonical URL:** [_wp16171](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16171.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16171.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16171.pdf.json)

---

### I. Introduction — purpose, scope, and main results
- Purpose: Apply the Synthetic Control Method (SCM) to investigate the impact of major product and labor reform episodes on income per capita in advanced economies over the past three decades.
- Case selection:
  - Initially selected nine cases (including two reform episodes in Australia, two in New Zealand and two in the Netherlands in the 1980s and 1990s, respectively, plus Denmark, Ireland in the 1990s, and Germany in the early 2000s).
  - Dropped Australia, New Zealand and the Netherlands in the 1980s due to poor pre-treatment fit.
  - Final sample: six case studies — Australia, New Zealand, Denmark, Ireland, and the Netherlands in the 1990s, and Germany in the early 2000s.
- Key high-level findings:
  - Structural reforms had positive impact on income per capita in four out of the six cases, compared to their synthetic control units.
  - The estimated impact on income per capita varies widely across cases but is generally sizeable.
  - Denmark: estimated reform impact is close to zero.
  - New Zealand: income per capita failed to catch up with its synthetic counterpart; this may partly reflect implementation under particularly weak macroeconomic conditions.
- Contributions relative to prior work:
  - Identifies heterogeneous, country-specific reform impacts rather than average treatment effects from cross-country OLS.
  - Uses data-driven construction of counterfactuals (synthetic controls) restricted to similar-income (advanced) economies.
  - Reports weights of each comparison unit contributing to the counterfactual, enabling quantitative and qualitative assessment.

### II. Synthetic Control Method (methodology and inference)
- SCM overview and advantages:
  - Selects the control unit as a linear combination of donor-pool countries that most closely match the treated unit in the pre-treatment period.
  - Matches pre-treatment outcomes and covariates by choosing weights that minimize pre-intervention RMSPE.
  - Allows case-by-case estimation of treatment effects, revealing heterogeneous impacts.
  - Reduces bias from time-varying unobserved confounders as pre-intervention period increases.
  - Makes explicit the contribution (weights) of each donor country to the synthetic control.
  - Enables placebo experiments for inference that avoid parametric assumptions about error structure.
- Limitations noted:
  - Does not address reverse causality (endogeneity) arising when reforms are driven by growth expectations not captured by unobservables.
  - Requires donor countries not to experience major idiosyncratic shocks during the sample period; in practice the authors only exclude other major reformers.
  - Does not exclude moderate or minor reforms in donor countries; cumulative minor reforms in donors could bias estimates downward (i.e., the estimates likely underestimate true impact).
- Technical implementation (selected details preserved exactly):
  - Treated country indexed as country 1 adopting reform at time T0 + 1.
  - Outcome of interest: log of real GDP per capita.
  - Linear factor model: y_it^NR = α_t + X_i β_t + Z_i δ_t + ε_it.
  - Synthetic control defined by a (J × 1) weight vector W = (w_2, ..., w_{J+1})′ with w_j ≥ 0 and sum of w_j = 1.
  - Optimal weights chosen by minimizing ∥X_1 − X_0 W∥_V, where V is symmetric, positive semi-definite and diagonal; minimization yields RMSPE-minimizing fit.
- Inference via placebo experiments:
  - Iteratively estimate placebo treatment effects by assuming each donor-pool country was treated in the same year as the treated country and constructing synthetic controls (omitting the treated country from donor pool).
  - Compare distribution of placebo effects to the treated-country effect to assess whether the observed effect is unusually large.
  - Limitation: placebo tests cannot rule out idiosyncratic shocks or other policy changes coinciding with treatment.

### III. Pre-treatment fit assessment and inference procedures
- Pre-treatment fit:
  - Root mean square prediction error (RMSPE) used to measure fit between treated and synthetic pre-treatment paths.
  - Authors use the “pre-treatment fit index” of Adhikari and Alm (2016), which normalizes RMSPE by a benchmark RMSPE (zero-fit model).
  - Interpretation:
    - Fit index of (0, x) indicates RMSPE equivalent to a zero-fit model with x percent difference per year.
    - Fit index = 0 → perfect fit.
    - Fit index = 1 → fit equal to zero-fit model.
    - Fit index > 1 → poor fit (e.g., synthetic with twice or half the GDP per capita of treated country is not a reasonable counterfactual).

### IV. Data and donor-pool selection rules
- Data sources:
  - Panel for 1960–2011 combining Penn World Table (PWT) version 8.1, World Development Indicators, and IMF World Economic Outlook.
  - Democracy from Polity IV; country classified as democratic if “polity2” > 0.
- Outcome and covariates:
  - Outcome: log of real GDP per capita.
  - Covariates: initial level of GDP per capita (log), stock of physical capital per capita (log), stock of human capital per capita, democracy, trade openness, and population growth.
- Sample window and calibration requirements:
  - Require 10 pre-treatment years to calibrate the synthetic unit and 5 post-treatment periods to evaluate impact.
  - Thus, for each case the sample window consists of 16 years.
- Donor-pool restrictions (applied exactly as in source):
  - Exclude countries that also had a major reform within five years before or after the treatment year.
  - Remove countries under major civil or ethnic violence during the sample period (data from Correlates of War project).
  - Exclude countries with missing values for the outcome variable in the sample window.
  - Restrict donor pool to countries in the same income group (Advanced economies vs Non-Advanced economies) at the year of treatment based on IMF classification (when early-year classification missing, label by first available year).

### V. Case studies — New Zealand (1990s) summary of reforms and context
- Reform motivation and timing:
  - Reverse relative decline vs OECD average; second wave of significant reforms starting around 1991.
- Major labor-market measures:
  - Employment Contracts Act of 1991: decentralized wage-bargaining; allowed negotiation of individual or collective contracts; parties could choose bargaining agents; no contract could require union membership or nonmembership.
  - Unemployment benefits: December 1990 package reduced benefits and tightened eligibility; wait period before entitlement increased from 6 to 26 weeks.
  - Active labor market policies: “Community Taskforce” (1991); beneficiaries could be required to work three days per week for an increment on benefits; July 1992 schemes expanded to include training assistance.
  - Pension system: July 1991 raise of standard eligibility age from 60 to 65, phased in between 1992 and 2001; benefits frozen for two years as opposed to inflation indexation.
- Product-market / trade measures and sector reforms:
  - Trade policy: March 1990 announced post-1992 tariff program in which most tariffs would fall to a maximum level of 10 percent between 1993 and [text ends here].
  - Public ownership and barriers to entry: privatizations in electricity, telecommunications, railways and hotels; telecommunications monopoly protections removed in 1989; taxi licensing liberalized in 1990; 1992 Electricity Act and July 1994 separation of transmission and generation; 1992 Gas Act and 1997 Information Disclosure Regulations.
- Macroeconomic context:
  - Reforms initiated during 1990-1991 recession.
  - Policy rates stood above 10 percent until 1991.
  - Real effective exchange rate had appreciated sharply in the late 1980s.
  - Fiscal policy became restrictive around 1991, contributing to an improvement in the cyclically-adjusted balance of over 3 percentage points between 1991 and 1993.
- Impact statistics:
  - Real GDP per capita grew by 2.5 percent over 1985-1990.
  - Real GDP per capita increased by 14.3 percent during 1991-1996—by 9 percent during 1990-1996, due to the large drop during the 1991 recession.
  - Average real GDP per capita across all advanced economies grew by 16.4 percent during 1985-1990, versus 9.4 percent over 1991-1996.
- Synthetic control analysis findings:
  - Relative to its synthetic control group, New Zealand’s GDP per capita dropped sharply in 1991 (the assumed treatment date).
  - GDP per capita level subsequently caught up to that of the synthetic control group but did not exceed it even five years after the reform.
  - Conclusion: unclear whether reforms gave a significant boost to living standards; timing during recession and restrictive macroeconomic policies may have muted effects.
- Synthetic control donor composition:
  - Synthetic New Zealand consists of Greece and the United States.

### VI. Case studies — Australia (1990s)
- Policy drivers and context:
  - Reforms aimed to promote growth and reduce unemployment after early-1990s recession; unemployment above 10 percent.
  - Macroeconomic policy eased in 1993: small-scale fiscal stimulus; central bank cut policy rate by 1 percentage point; exchange rate depreciated.
  - Overall policy stance mostly restrictive: fiscal consolidation reduced the cyclically-adjusted primary deficit from about 4 percent of potential GDP in 1993 to virtually zero in 1997; policy rate rose from 4.75 percent in 1993 to 7.5 percent in 1995.
- Labor market reforms:
  - Transition toward firm-level bargaining; 1992 Industrial Relations Legislation Amendment Act removed “public interest” test.
  - March 30, 1994: 1993 Industrial Relations Reform Act allowed registering collective non-union agreements.
  - Active labor market policies: Working Nation (May 1994); Job Compact; 1998 restructuring of Commonwealth Employment Service; lowered withdrawal rate for benefits from 100% to 70% of each additional dollar earned.
- Product market liberalization and state ownership changes:
  - August 1994 COAG package on access to essential facilities; July 1993 COAG agreement for competitive bulk electricity market from July 1995.
  - Privatizations: Australian National business units mid 1997; Qantas privatization March 1993–late 1995; government reduced share in Commonwealth Bank from 70 to 51 percent in 1994.
  - Competition institutions: Australian Competition Commission (ACC) and National Competition Council (NCC); Competition Policy Reform Act (1995) took effect in November 1995.
- Impact statistics:
  - Income per capita grew by 15.6 percent over the five-year post-reform period (1994-1999), compared to 5.5 percent in the five years prior to the reforms (1988-1993).
  - Advanced economies experienced a pick-up from 8.4 to 12.8 percent between these periods.
- Synthetic control analysis findings:
  - GDP per capita levels were 1.4 percent higher in Australia than in the synthetic control group five years after the reform, while levels were similar in 1994.
  - Placebo experiments were not conclusive: half of the fake experiments in potential controls yield a treatment effect below that of Australia; small number of placebo countries makes statistical inference difficult.
  - Tentative conclusion: Australia’s growth might have benefited from the product and labor market reforms during the second half of the 1990s; strong performance throughout the 2000s could partly reflect delayed effects.
- Synthetic control donor composition:
  - Synthetic Australia consists of Greece and the United States.

### VII. Case studies — Netherlands (1990s)
- Reform drivers and timing:
  - Early 1990s downturn raised unemployment from about 6 percent in 1991 to 10 percent in 1994.
  - Measures to enhance wage flexibility, boost job search and creation incentives, reduce state control, and open major industries to competition.
- Major labor-market and product-market measures (selected):
  - From 1994, administrative extensions of labor agreements no longer applied to the lower end of the market.
  - October 1995: agreement to reduce gap between legal minimum wage and collective agreement minima; “opening clauses” promoted.
  - Active labor market changes: TBA Act (August 1993); January 1995 programs to increase temporary jobs; March 1996 abolition of public sickness benefit scheme; August 1996 ban on unemployment benefits where unemployment deemed voluntary; early 1997 information-sharing law; January 1999 Flexibility and Security Law.
  - Product market liberalization: 1993 EU air transport directive implementation; 1994–1995 KPN stake sales; railway reform from 1995; Shop Hours Act of 1996; 1997 new telephone operators; late 1990s opening of electricity, telecommunications, pharmaceuticals and gas; 1999 Gas Act and 1999 opening of electricity market.
  - Competition and administrative burden: 1993 decrees under 1956 Economic Competition Act; January 1996 Liberalized Establishment Law reduced number of establishment licenses from 88 to 8.
- Macroeconomic context:
  - Monetary easing with policy rate cuts of over 400 basis points between 1992 and 1994; easing continued afterward; fiscal consolidation of early 1990s halted in 1994-1995 with small weakening of cyclically-adjusted balance.
- Impact statistics:
  - Real GDP per capita grew by 18.3 percent over 1994-1999, versus 10.8 percent during 1988-1993.
  - Real GDP per capita in the Netherlands was very close to that of its synthetic control group in 1994; after reforms a gap opened reaching 6 percent in 1999.
  - Reported aggregate pick-up in advanced economies: 12.8 percent during 1994-1999.
- Synthetic control composition:
  - Synthetic Netherlands consists of Austria, Belgium, Greece, Switzerland, and the United States.
- Inference:
  - Placebo test finds that over half of the control countries have a smaller estimated treatment effect than that of the Netherlands—statistical inference hampered by small number of units.

### VIII. Case studies — Denmark (1990s)
- Reform components:
  - Active labor market policy: mandatory participation after extended unemployment; young people below 25 required to take 18 months training at 50 percent benefit after six months of unemployment (policy entered into force April 1996, fully phased in April 1997).
  - Unemployment benefit duration/eligibility: maximum benefit period limited to seven years in 1994-95 (extendable by two years via paid-leave); from 1996 maximum effective benefit period reduced to five years with last three involving full-time activation; eligibility tightened to require 52 weeks of non-subsidized work over the last three years.
  - Tax reforms: marginal tax rates announced to be gradually reduced by 8 to 14 percentage points over 1994-1998; combined marginal rate lowered by 6 to 9 percentage points for most employees.
  - Product market deregulation and privatizations: sales of shares in state companies (including 25 percent shares in Copenhagen Airport and the Postal Giro and 49 per cent of the shares in TeleDanmark in 1993); deregulation/liberalization in domestic airline industry (1995), telecommunications (1996), rail transport framework (1997), and competition in electricity (1999 Energy Supply Act).
- Macroeconomic context:
  - Reforms initiated as growth was starting to recover from the early 1990s recession.
  - Monetary policy accommodative after 1993; policy rates reduced drastically alongside Germany’s over 1993-1994.
  - New government (1993) pursued medium-term strategy to eliminate general government deficit by 1997; short-term fiscal stimulus of about 1.5 percentage points of GDP implemented with later consolidation.
- Impact statistics:
  - Real GDP per capita grew by 12.7 percent in 1994-1999, compared to 4.5 percent in 1988-1993.
  - GDP per capita gap between Denmark and its synthetic equivalent remained roughly constant over the 10 years around the treatment date (1994).
  - Synthetic control analysis suggests an insignificant impact of the reforms on GDP per capita in the first few years after the start of the episode.
  - Donor units (fit index): 3/13 (0.03). Placebo units (p-value): 12 (0.33).

### IX. Case studies — Ireland (1990s)
- Reform components:
  - Unemployment benefits: replacement rates lowered—average replacement rate reduced from 42 to 35 percent for a single person and from 70 to 64 percent for a married person with two children.
  - Back to Work Allowance Scheme allowed long-term unemployed to accept jobs while partly retaining benefits; 1995 Budget further reduced replacement ratios.
  - Active labor market policies scaled up from 1994; Community Employment Programme doubled long-term unemployed participants (participants about 3 per cent of total employment); 1997 measures included job subsidies and scope to combine wage income with unemployment benefit receipt.
  - Collective bargaining: two centralized wage agreements (Programme for Competitiveness and Work 1993-1996; Partnership 2000 for 1997-1999) and sharp cut in labor income taxation.
  - Product market reforms: 1993 EU Air Transport Liberalization Directive (from 1994); gas sector entry lowered; 1998 legislation to partially deregulate electricity; telecommunications liberalized in 1998; 1994 amendments strengthened Competition Authority; privatizations and restructuring of state-owned enterprises.
- Macroeconomic context:
  - Reform carried out under robust economic growth; central bank lowered policy rates sharply in 1993-1994 after a 10 percent devaluation within the European Monetary System; monetary policy tightened in 1995-1996.
  - Fiscal policy supportive in 1994; cyclically-adjusted balance weakened over 1994-1996 despite strong revenue growth.
- Impact statistics:
  - Real GDP per capita grew by about 21 percent over 1988-1993, whereas it rose by more than 54 percent during 1994-1999.
  - Synthetic control results: real GDP per capita in Ireland was 6.8 percent higher than in the synthetic unit in 1994, and 34 percent higher in 1999.
  - Placebo experiment: estimated impact of reforms in Ireland larger than in all placebo units.
  - Donor units (fit index): 3/13 (0.03). Placebo units (p-value): 12 (0.00).
- Note: Other exogenous factors likely contributed to growth, including a boom in FDI inflows from the United States and possible lagged effects from late-1980s labor market reforms.

### X. Case studies — Germany (Hartz reforms in the 2000s)
- Reform components:
  - Hartz I-III implemented in 2003-2004; Hartz IV implemented in 2005.
  - Hartz I: creating new types of employment opportunities; Hartz II: additional wage subsidies; Hartz III: restructuring the Federal Employment Agency.
  - Hartz IV: significant cut in unemployment benefits for the long-term unemployed (merged with lower-level welfare benefits); shortening of eligibility period for unemployment insurance benefits for older employees; tightening of job search requirements.
  - Accompanying decentralization of bargaining from industry/region to firm/individual levels.
- Macroeconomic context:
  - Growth weak and unemployment rising at reform start.
  - Monetary policy accommodative: ECB cut main refinancing rate by 125 basis points over 2002-2003 and left it at 2 percent during 2003-2005.
  - Fiscal policy: priority on gradual fiscal consolidation and public sector expenditure reforms.
- Impact statistics:
  - Considering 2003 as start date, real GDP per capita in Germany grew by about 11 percent in 2003-2008, compared to 8.3 percent in 1997-2002.
  - Growth of average real GDP per capita in all advanced economies slowed from 10.6 percent in the pre-reform period to 8.7 percent in the post-reform period.
  - Synthetic control analysis: GDP per capita gap between Germany and its synthetic counterpart rose from 4.2 percent in 2003 to 9.3 percent in 2008.
  - Placebo experiment: 18 percent of placebo units showed a larger estimated impact of reform than Germany.
  - Placebo units (p-value): 22 (0.18).

### XI. Cross-case methodological findings, statistical significance, and limitations
- Method and sample:
  - Approach: Synthetic control method used to construct counterfactuals, taking the starting year of the reform wave as the treatment date.
  - Sample: Six individual cases analyzed—New Zealand, Australia, The Netherlands, Denmark, Ireland (1990s product and labor market reform packages) and Germany (Hartz reforms in the 2000s).
- Statistical significance:
  - In four out of six cases there is some evidence of a pick-up in GDP per capita relative to the counterfactual within five years after the reform start.
  - Exceptions: Denmark (no estimated effect relative to the counterfactual, although GDP per capita growth did pick up) and New Zealand (initial pick-up smaller than the no-reform counterfactual with some later catch-up).
  - With the exception of Ireland, placebo experiments do not permit claiming statistical significance of the results.
- Limitations and caveats emphasized by authors:
  - SCM cannot fully rule out other major idiosyncratic shocks around the treatment date.
  - SCM cannot fully rule out reverse causality when reforms are driven by poor prospects not captured by unobservables.
  - Major reform waves are often phased in over several years, complicating clean treatment identification.
  - Minor reforms in control countries may cumulatively affect outcomes; these limitations imply estimates are more likely to underestimate, rather than overestimate, the true impact of major structural reform waves.

### XII. Tentative policy conclusions
- Major reform packages are typically implemented gradually over several years and often take time to pay off, making it difficult to identify systematically large effects in their immediate aftermath.
- Supportive macroeconomic policies seem to enhance the short-term pay-off from reforms. New Zealand—where macroeconomic support was lacking when major reforms were implemented in the early 1990s—illustrates this point.

*Source: _wp16171 - References (International Monetary Fund Working Paper excerpt).*

### References .............................................................................................................

### _wp16171 - References

### I. Introduction — purpose, scope, and main results
- Purpose: Apply the Synthetic Control Method (SCM) to investigate the impact of major product and labor reform episodes on income per capita in advanced economies over the past three decades.
- Case selection:
  - Initially selected nine cases (including two reform episodes in Australia, two in New Zealand and two in the Netherlands in the 1980s and 1990s, respectively, plus Denmark, Ireland in the 1990s, and Germany in the early 2000s).
  - Dropped Australia, New Zealand and the Netherlands in the 1980s due to poor pre-treatment fit.
  - Final sample: six case studies — Australia, New Zealand, Denmark, Ireland, and the Netherlands in the 1990s, and Germany in the early 2000s.
- Key high-level findings:
  - Structural reforms had positive impact on income per capita in four out of the six cases, compared to their synthetic control units.
  - The estimated impact on income per capita varies widely across cases but is generally sizeable.
  - Denmark: estimated reform impact is close to zero.
  - New Zealand: income per capita failed to catch up with its synthetic counterpart; this may partly reflect implementation under particularly weak macroeconomic conditions.
- Contributions relative to prior work:
  - Identifies heterogeneous, country-specific reform impacts rather than average treatment effects from cross-country OLS.
  - Uses data-driven construction of counterfactuals (synthetic controls) restricted to similar-income (advanced) economies.
  - Reports weights of each comparison unit contributing to the counterfactual, enabling quantitative and qualitative assessment.

### II. Synthetic Control Method (methodology and inference)
- SCM overview and advantages:
  - Selects the control unit as a linear combination of donor-pool countries that most closely match the treated unit in the pre-treatment period.
  - Matches pre-treatment outcomes and covariates by choosing weights that minimize pre-intervention RMSPE.
  - Allows case-by-case estimation of treatment effects, revealing heterogeneous impacts.
  - Reduces bias from time-varying unobserved confounders as pre-intervention period increases.
  - Makes explicit the contribution (weights) of each donor country to the synthetic control.
  - Enables placebo experiments for inference that avoid parametric assumptions about error structure.
- Limitations noted:
  - Does not address reverse causality (endogeneity) arising when reforms are driven by growth expectations not captured by unobservables.
  - Requires donor countries not to experience major idiosyncratic shocks during the sample period; in practice the authors only exclude other major reformers.
  - Does not exclude moderate or minor reforms in donor countries; cumulative minor reforms in donors could bias estimates downward (i.e., the estimates likely underestimate true impact).
- Technical implementation (selected details preserved exactly):
  - Treated country indexed as country 1 adopting reform at time T0 + 1.
  - Outcome of interest: log of real GDP per capita.
  - Linear factor model: y_it^NR = α_t + X_i β_t + Z_i δ_t + ε_it (notation and full formulae provided in source).
  - Synthetic control defined by a (J × 1) weight vector W = (w_2, ..., w_{J+1})′ with w_j ≥ 0 and sum of w_j = 1.
  - Optimal weights chosen by minimizing ∥X_1 − X_0 W∥_V, where V is symmetric, positive semi-definite and diagonal; minimization yields RMSPE-minimizing fit.

### III. Pre-treatment fit assessment and inference procedures
- Pre-treatment fit:
  - Root mean square prediction error (RMSPE) used to measure fit between treated and synthetic pre-treatment paths.
  - Authors use the “pre-treatment fit index” of Adhikari and Alm (2016), which normalizes RMSPE by a benchmark RMSPE (zero-fit model).
  - Interpretation:
    - Fit index of (0, x) indicates RMSPE equivalent to a zero-fit model with x percent difference per year.
    - Fit index = 0 → perfect fit.
    - Fit index = 1 → fit equal to zero-fit model.
    - Fit index > 1 → poor fit (e.g., synthetic with twice or half the GDP per capita of treated country is not a reasonable counterfactual).
- Inference via placebo experiments:
  - Iteratively estimate placebo treatment effects by assuming each donor-pool country was treated in the same year as the treated country and constructing synthetic controls (omitting the treated country from donor pool).
  - Compare distribution of placebo effects to the treated-country effect to assess whether the observed effect is unusually large.
  - Limitation: placebo tests cannot rule out idiosyncratic shocks or other policy changes coinciding with treatment.

### IV. Data and donor-pool selection rules
- Data sources:
  - Panel for 1960–2011 combining Penn World Table (PWT) version 8.1, World Development Indicators, and IMF World Economic Outlook.
  - Democracy from Polity IV; country classified as democratic if “polity2” > 0.
- Outcome and covariates:
  - Outcome: log of real GDP per capita.
  - Covariates: initial level of GDP per capita (log), stock of physical capital per capita (log), stock of human capital per capita, democracy, trade openness, and population growth.
- Sample window and calibration requirements:
  - Require 10 pre-treatment years to calibrate the synthetic unit and 5 post-treatment periods to evaluate impact.
  - Thus, for each case the sample window consists of 16 years.
- Donor-pool restrictions (applied exactly as in source):
  - Exclude countries that also had a major reform within five years before or after the treatment year.
  - Remove countries under major civil or ethnic violence during the sample period (data from Correlates of War project).
  - Exclude countries with missing values for the outcome variable in the sample window.
  - Restrict donor pool to countries in the same income group (Advanced economies vs Non-Advanced economies) at the year of treatment based on IMF classification (when early-year classification missing, label by first available year).

### V. Case studies — New Zealand (1990s) summary of reforms and context
- Reform motivation: reverse relative decline in New Zealand’s economic performance vs OECD average and address distortive economic policies (high tariffs, import licensing/quotas, complex regulations, subsidies).
- Timing: second wave of significant reforms starting around 1991, including measures to enhance labor market flexibility and strengthen competition in product markets (domestic deregulation and further reduction in trade protection).
- Major labor-market measures described:
  - Collective wage bargaining: Employment Contracts Act of 1991 introduced decentralized wage-bargaining and allowed negotiation of individual or collective contracts; parties could choose bargaining agents; no contract could require union membership or nonmembership.
  - Unemployment benefits: December 1990 package reduced benefits and tightened eligibility; wait period before entitlement increased from 6 to 26 weeks.
  - Active labor market policies: “Community Taskforce” (1991) provided work experience and work-testing; beneficiaries could be required to work three days per week for an increment on benefits; in July 1992 schemes expanded to include training assistance focused on long-term unemployed.
  - Pension system: In July 1991 standard eligibility age for retirement benefits raised from 60 to 65, phased in gradually between 1992 and 2001; benefits were frozen for two years as opposed to inflation indexation.
- Product-market / trade measures:
  - Trade policy: In March 1990 government announced a post-1992 tariff program in which most tariffs would fall to a maximum level of 10 percent between 1993 and [text ends here].

### VI. Limitations and caveats emphasized by authors
- SCM cannot fully rule out:
  - Other major idiosyncratic shocks around the treatment date.
  - Reverse causality when reforms are driven by poor prospects not captured by unobservables.
  - Challenges in finding a clean treatment since major reform waves are often phased in over several years.
  - Minor reforms in control countries that may cumulatively affect outcomes.
- Directional implication: these limitations imply estimates are more likely to underestimate, rather than overestimate, the true impact of major structural reform waves.

*Source: _wp16171 - References*

### 1996. Further programs of tariff reductions were announced and implemented in the

### _wp16171 - 1996. Further programs of tariff reductions were announced and implemented in the

### Public ownership and barriers to entry (New Zealand, early 1990s)
- Public ownership
  - Early 1990s saw a wave of full or partial privatization of state-owned enterprises, including in electricity, telecommunications, railways and hotels.
- Barriers to entry and sector-specific reforms
  - All statutory protections of the telecommunications monopoly were removed and other firms were able to provide network services in competition with the incumbent (Telecom) in the course of 1989.
  - Taxi licensing was liberalized by the Ministry of Transport in 1990: fares allowed to fluctuate based on market conditions and geographic restrictions on number of licenses were removed.
  - The 1992 Electricity Act introduced potential competition between energy retailers by allowing sale to any consumer using existing distribution lines; reform implemented in two steps in April 1993 (normal consumers) and April 1994 (large consumers).
  - In July 1994 the government separated electricity transmission (Transpower) from generation (Electricorp).
  - The 1992 Gas Act and 1997 Information Disclosure Regulations fostered competition in the gas market through termination of exclusive franchise arrangements, information disclosure regulation, and removal of price controls.

### Macroeconomic context (New Zealand)
- The wave of reforms was initiated during a downturn—the 1990-1991 recession—to help regain external competitiveness.
- Macroeconomic policy was restrictive:
  - Policy rates stood above 10 percent until 1991 as the Reserve Bank sought rapid disinflation under the 1989 Reserve Bank Act.
  - The real effective exchange rate had appreciated sharply in the late 1980s, tightening financial conditions.
  - Fiscal policy became restrictive around 1991, contributing to an improvement in the cyclically-adjusted balance of over 3 percentage points between 1991 and 1993.

### Impact of the reforms (New Zealand)
- Growth statistics
  - Real GDP per capita grew by 2.5 percent over the five-year period 1985-1990.
  - Real GDP per capita increased by 14.3 percent during 1991-1996—by 9 percent during 1990-1996, due to the large drop during the 1991 recession.
  - Average real GDP per capita across all advanced economies grew by 16.4 percent during 1985-1990, versus 9.4 percent over 1991-1996.
- Synthetic control analysis findings
  - Relative to its synthetic control group, New Zealand’s GDP per capita dropped sharply in 1991 (the assumed treatment date).
  - New Zealand’s GDP per capita level subsequently caught up to that of the synthetic control group but did not exceed it even five years after the reform.
  - Conclusion: on the basis of the exercise, it is unclear whether the reforms gave a significant boost to living standards; timing during recession and restrictive macroeconomic policies may have muted effects.
- Literature evidence and interpretations
  - Evans and others (1996): inadequate macroeconomic policy mix and sequencing partly explain why reform was not as successful as expected, but emphasize “broadly positive macroeconomic outcomes”.
  - Glyn (2005): New Zealand’s poor productivity performance in the 1990s relative to Australia and Ireland is not indicative of a strong positive effect of reforms.
  - McCann (2003); Boulhol, de Serres and Molnar (2008): lack of agglomeration effects and geographical distance from key markets may have prevented major efficiency gains.
  - New Zealand Treasury (Galt 2000): growth benefits from reforms might be slow to appear and not fully measured, though it is unclear why measurement would have worsened after reforms relative to comparators.
- Note on synthetic control donor composition
  - Synthetic New Zealand consists of Greece and the United States.

*Source: _wp16171 - 1996. Further programs of tariff reductions were announced and implemented in the*

---

### Labor and product market reforms in Australia in the 1990s

### Policy drivers and context (Australia)
- The Australian economy entered recession in the early 1990s with slow growth and unemployment above 10 percent.
- Reform objective: promote growth and reduce unemployment via a new wave of major labor and product market reforms in the mid-1990s.

### Labor market reforms (major elements)
- Collective bargaining
  - Transition away from centralized bargaining toward firm-level bargaining.
  - 1992 Industrial Relations Legislation Amendment Act removed requirement for certified enterprise-level agreements to meet a “public interest” test.
  - March 30, 1994: 1993 Industrial Relations Reform Act allowed registering collective non-union agreements for the first time since 1904; incentives and penalties for single-employer bargaining; non-union agreements could be formally registered.
  - Outcome: union density declined significantly while growth in non-union collective agreements picked up.
- Active labor market policies
  - Working Nation program launched in May 1994; Job Compact focused on persons on unemployment support for at least 18 months offering job placement (including wage subsidies) for 6 to 12 months and intensified job-search assistance afterward.
  - Strengthened labor market assistance for non-job-compact workers, emphasis on those at risk of long-term unemployment.
  - 1998: restructuring of Commonwealth Employment Service (CES) and opening provision of case management services to competition.
  - Incentive changes: lowered withdrawal rate for benefits from 100% to 70% of each additional dollar earned.
  - Youth Training Initiative: case management assistance and subsidies for unemployed under age 18.

### Strengthening competition in product markets (Australia)
- Barriers to entry and access regimes
  - August 1994 COAG package: regime for access to essential facilities (electricity grids, gas pipelines, airports, rail networks, postal delivery services, communication channels, seaports); agreed principles on structural reform of public monopolies; competitive neutrality and review of regulations restricting competition.
  - July 1993 COAG agreement to create competitive market for bulk electricity in southern and eastern Australia from July 1995.
  - June 1995: access price system for Commonwealth’s rail track announced; July 1995: competitive electricity market commencement agreed.
  - June 1994: more competition in gas transmission when Moomba-Sydney Pipeline was sold with provisions to encourage competition.
  - July 1997: full and open competition began in telecommunications.
- State ownership changes
  - Privatizations: business units of Australian National in mid 1997; Qantas privatization started March 1993 and completed late 1995.
  - Government reduced share in Commonwealth Bank of Australia from 70 to 51 percent in 1994.
- Competition policy institutions and laws
  - August 1994 COAG package introduced price surveillance systems and established Australian Competition Commission (ACC) and National Competition Council (NCC).
  - April 1995: COAG agreed to implement National Competition Policy (NCP).
  - Commonwealth’s Competition Policy Reform Act (1995) extended competitive conduct rules to unincorporated and government business enterprises and amended Price Surveillance Act; took effect in November 1995.

### Macroeconomic context (Australia)
- Reforms started after an early-1990s recession; macro policy eased in 1993:
  - Small-scale fiscal stimulus to support recovery.
  - Central bank cut policy rate by 1 percentage point.
  - Exchange rate depreciated.
- Overall policy stance mostly restrictive during reform period:
  - Fiscal consolidation reduced the cyclically-adjusted primary deficit from about 4 percent of potential GDP in 1993 to virtually zero in 1997.
  - Monetary policy prioritized low inflation; policy rate rose from 4.75 percent in 1993 to 7.5 percent in 1995.

### Impact of the reforms (Australia)
- Growth statistics
  - Income per capita grew by 15.6 percent over the five-year post-reform period (1994-1999), compared to 5.5 percent in the five years prior to the reforms (1988-1993).
  - Advanced economies experienced a pick-up from 8.4 to 12.8 percent between these periods.
- Synthetic control analysis findings
  - GDP per capita levels were 1.4 percent higher in Australia than in the synthetic control group five years after the reform, while levels were similar in 1994.
  - Placebo experiments were not conclusive: half of the fake experiments in potential controls yield a treatment effect below that of Australia; small number of placebo countries makes statistical inference difficult.
  - Tentative conclusion: Australia’s growth might have benefited from the product and labor market reforms during the second half of the 1990s; strong performance throughout the 2000s could partly reflect delayed effects.
- Supporting and contrasting literature
  - Parham (2002): reforms played a key part in the productivity surge of the 1990s.
  - IMF (2015): finds positive effects of the 1998-2001 wave of fiscal reforms using the synthetic control method.
  - Quiggin (2001, 2004): more skeptical.

- Note on synthetic control donor composition
  - Synthetic Australia consists of Greece and the United States.

*Source: _wp16171 - 1996. Further programs of tariff reductions were announced and implemented in the*

---

### Labor and product market reforms in the Netherlands in the 1990s

### Reform drivers (Netherlands)
- Early 1990s downturn raised unemployment from about 6 percent in 1991 to 10 percent in 1994.
- Government measures aimed to enhance wage flexibility, boost job search and creation incentives, reduce state control, and open major industries (gas, electricity, telecommunications) to competition.

### Deregulation of the labor market (major elements)
- Minimum wage and collective agreements
  - From 1994, administrative extensions of labor agreements no longer applied to the lower end of the market.
  - October 1995: employers and workers’ organizations agreed to reduce the gap between the legal minimum wage and minimum wages set in collective labor agreements.
  - Promotion of “opening clauses” allowed firms under certain circumstances to negotiate pay below collective-contract minima.
- Active labor market policies and benefit conditionality
  - TBA Act (Restriction of Claims on the Disability Benefit Regulations) of August 1993 reduced generosity of disability scheme and introduced more stringent conditions for access.
  - January 1995: three programs introduced to increase temporary jobs for long-term unemployed.
  - June 1995: Minister of Social Affairs approved 57 subsidized job programs in the private sector targeted at long-term unemployed.
  - March 1996: public sickness benefit scheme abolished to reduce sickness-related absence.
  - August 1996: law banned unemployment benefits where unemployment deemed voluntary; refusal to accept a suitable job offer could result in complete and permanent withdrawal of unemployment benefit.
  - Early 1997: law obliging Public Employment Services to inform unemployment benefit agencies of insufficient job search efforts by claimants.
- Tax wedges and hiring incentives
  - January 1996: special measure to reduce employers’ social contributions for workers with wages less than 115 percent of legal minimum wage.
  - Costs of hiring a long-term unemployed person at wages up to 130 percent of legal minimum were reduced by about 13 percent.
- Employment protection legislation
  - December 1995: decision to shorten dismissal procedures; employer could dismiss at same time or before seeking permission from director of Public Employment Service.
  - Early 1998: new law on fixed-term contracts, probation periods and dismissal procedures removed restrictions on renewal of fixed-term contracts and slightly eased dismissal procedures for permanent workers.
  - January 1999: Flexibility and Security Law promoted flexible working contracts and streamlined dismissal procedures.

### Product market liberalization (major elements)
- Barriers to entry and sector openings
  - 1993: implementation of EU Directive deregulating air transport.
  - Postal services: government sold 30 percent stake in KPN in 1994 and another 22 percent stake in 1995, reducing government share to 48 percent.
  - Railway reform initiated in 1995 enabled competition; new operators appeared from 1996.
  - Retail trade: Shop Hours Act of 1996 liberalized shop opening hours.
  - 1997: two new national telephone operators allowed to enter market; additional mobile telephony frequencies planned for auction after late 1995 auctions.
  - Late 1990s: wave of reforms opened up electricity, telecommunications, pharmaceuticals and gas.
  - 1998: Postal Directive of the European Commission implemented to open some segments of mail services.
  - 1999: Dutch electricity market opened to competition—licenses no longer required for production and large users allowed to choose among suppliers.
  - 1999 Gas Act started gradual liberalization in gas distribution.
- Competition policy and administrative burden reduction
  - Three decrees under 1956 Economic Competition Act introduced; first decree prohibited horizontal price agreements from July 1993 with effective implementation over 1994 (government refused 12 requests for exceptions).
  - September 1993: amendments to Economic Competition Act broadened scope to cover professional services and improved legal action possibilities against informal restrictive practices.
  - January 1996: Liberalized Establishment Law reduced number of establishment licenses from 88 to 8 to reduce barriers to entry and lower administrative costs for startups.

### Macroeconomic context (Netherlands)
- Reforms introduced against background of high unemployment after early-1990s recession.
- Macroeconomic policy mix turned accommodative around start of reform wave in 1994:
  - Strong confidence in Dutch guilder during 1993 ERM crisis allowed lowest interest rates in EU and policy rate cuts of over 400 basis points between 1992 and 1994 that continued afterward.
  - Easing of monetary conditions was substantial, dampened by appreciation of the real effective exchange rate.
  - Fiscal consolidation of early 1990s came de facto to a halt in 1994-1995 with a small weakening of the cyclically-adjusted balance.

### Impact of the reforms (Netherlands)
- Growth statistics
  - Real GDP per capita grew by 18.3 percent over 1994-1999, versus 10.8 percent during 1988-1993 (pre-reform period).
  - The pick-up in average real GDP per capita across advanced economies was milder—from 8.4 percent over 1988-1993 to (text truncates thereafter in source).

*Source: _wp16171 - 1996. Further programs of tariff reductions were announced and implemented in the*

### 12.8 percent during 1994-1999.

### _wp16171 - 12.8 percent during 1994-1999.

### Synthetic control composition
- Synthetic Netherlands consists of Austria, Belgium, Greece, Switzerland, and the United States.

### GDP per capita trajectory and treatment effect
- Real GDP per capita in the Netherlands was very close to that of its synthetic control group in 1994.
- After the start of structural reform episode, a gap opened up that reached 6 percent in 1999.
- 12.8 percent during 1994-1999.

### Placebo test and inference
- The placebo test finds that over half of the control countries have a smaller estimated treatment effect than that of the Netherlands—although, again, it is difficult to draw strong

*Source: _wp16171 - 12.8 percent during 1994-1999.*

### conclusions from such tests given the small number of units. Overall, there is tentative

### V.   CONCLUDING REMARKS

### Country case findings and impacts
- Netherlands
  - Tentative evidence that the major labor and product market reforms carried out in the Netherlands starting from the mid-1990s had some positive impact on GDP per capita growth in the five years after the start of the episode.
- Denmark (reforms in the 1990s)
  - Reform components:
    - Active labor market policy: mandatory participation in training/education or taking up suitable job offers after four years of unemployment (1994); compulsory full-time participation after two years of unemployment (implemented 1996, phased in by April 1997); young people below 25 on benefits had to take 18 months training at 50 percent benefit after six months of unemployment (new policy entered into force April 1996, fully phased in April 1997).
    - Unemployment benefit duration and eligibility: maximum benefit period limited to seven years in 1994-95 (extendable by two years via paid-leave); from 1996 maximum effective benefit period reduced to five years with last three involving full-time activation; eligibility tightened to require 52 weeks of non-subsidized work (formerly 26 weeks) over the last three years; paid leave for education could no longer extend the benefit period.
    - Tax reforms: marginal tax rates on personal incomes announced to be gradually reduced by 8 to 14 percentage points over 1994-1998; payroll taxes introduced to finance labor-market policy measures; combined marginal rate of personal income tax and social security contributions was lowered by 6 to 9 percentage points for most employees.
    - Product market deregulation: transformation of state enterprises into limited liability companies; sales of shares in state companies (including 25 percent shares in Copenhagen Airport and the Postal Giro and 49 per cent of the shares in TeleDanmark in 1993); retail trade opening hours liberalized March 1994 (unregulated from 8 a.m. Monday to 5 p.m. Saturday; some low-turnover shops allowed Sunday opening); deregulation/liberalization in domestic airline industry (1995), telecommunications (1996), rail transport framework (1997), and competition in electricity (1999 Energy Supply Act).
  - Macroeconomic context:
    - Reforms initiated as growth was starting to recover from the early 1990s recession.
    - Monetary policy accommodative after 1993 stress period; policy rates reduced drastically alongside Germany’s over 1993-1994.
    - New government (1993) pursued medium-term strategy to improve public finances with operational target of eliminating the general government deficit by 1997; short-term fiscal stimulus of about 1.5 percentage points of GDP implemented with later consolidation; cyclically-adjusted general government deficit widened in 1995 before gradually shrinking.
  - Impact statistics:
    - Real GDP per capita grew by 12.7 percent in the five years after the start of the reform episode (1994-1999), compared to 4.5 percent in the pre-reform period (1988-1993).
    - GDP per capita growth also picked up in advanced economies from 8.4 percent over 1988-1993 to 12.8 percent over 1994-1999.
    - Synthetic control analysis: suggests an insignificant impact of the reforms on GDP per capita in the first few years after the start of the reform episode; the GDP per capita gap between Denmark and its synthetic equivalent remained roughly constant over the 10 years around the treatment date (1994).
    - Donor units (fit index): 3/13 (0.03). Placebo units (p-value): 12 (0.33).
- Ireland (reforms in the 1990s)
  - Reform components:
    - Unemployment benefits: replacement rates lowered—average replacement rate reduced from 42 to 35 percent for a single person and from 70 to 64 percent for a married person with two children; Back to Work Allowance Scheme allowed long-term unemployed to accept jobs while partly retaining benefits; 1995 Budget further reduced replacement ratios and narrowed gap between child benefits for unemployed and employees.
    - Active labor market policies: from 1994 scaled up active policies; Community Employment Programme doubled long-term unemployed participants in activation activities (including part-time and public-sector jobs and training); participants accounted for about 3 per cent of total employment; 1997 measures included job subsidies and scope to combine wage income with unemployment benefit receipt; fourteen local employment services offices established.
    - Collective bargaining and labor taxation: two centralized wage agreements (Programme for Competitiveness and Work 1993-1996; Partnership 2000 for 1997-1999) and sharp cut in labor income taxation.
    - Product market reforms: implementation of 1993 EU Air Transport Liberalization Directive (from 1994); barriers to entry lowered in gas sector; 1998 legislation to partially deregulate electricity supply; telecommunications liberalized in 1998 with independent regulator; 1994 amendments strengthened Competition Authority enforcement; privatizations (Irish Life Plc. 1995; Irish Steel Ltd. 1996; 20 percent sale in Telecom Eireann end-1996) and restructuring of loss-making state-owned enterprises.
  - Macroeconomic context:
    - Reform carried out under robust economic growth; central bank lowered policy rates sharply in 1993-1994 after a 10 percent devaluation within the European Monetary System; monetary policy tightened in 1995-1996 amid rapid credit growth and rising asset prices.
    - Fiscal policy supportive in 1994 (lowered personal income tax on low-income earners, increased social welfare benefits, increased expenditure on community employment schemes); cyclically-adjusted balance weakened over 1994-1996 despite strong revenue growth.
  - Impact statistics:
    - Real GDP per capita grew by about 21 percent over the five years (1988-1993) prior to the reforms, whereas it rose by more than 54 percent during 1994-1999.
    - Synthetic control results: real GDP per capita in Ireland was 6.8 percent higher than in the synthetic unit in 1994, and 34 percent higher in 1999.
    - Placebo experiment: estimated impact of reforms in Ireland larger than in all placebo units.
    - Donor units (fit index): 3/13 (0.03). Placebo units (p-value): 12 (0.00).
  - Note: Other exogenous factors likely contributed to growth, including a boom in FDI inflows from the United States (owing to low corporation tax), and possible lagged effects from late-1980s labor market reforms.
- Germany (Hartz reforms in the 2000s)
  - Reform components:
    - Hartz I-III implemented in 2003-2004; Hartz IV implemented in 2005.
    - Hartz I: creating new types of employment opportunities; Hartz II: additional wage subsidies; Hartz III: restructuring the Federal Employment Agency.
    - Hartz IV: significant cut in unemployment benefits for the long-term unemployed (merged with lower-level welfare benefits); shortening of eligibility period for unemployment insurance benefits for older employees; tightening of job search requirements.
    - Accompanying decentralization of bargaining from industry/region to firm/individual levels.
  - Macroeconomic context:
    - Growth weak and unemployment rising at reform start.
    - Monetary policy accommodative: ECB cut main refinancing rate by 125 basis points over 2002-2003 and left it at 2 percent during 2003-2005; real short-term interest rates were higher in Germany relative to other euro-area countries due to comparatively lower inflation.
    - Fiscal policy: priority on gradual fiscal consolidation and public sector expenditure reforms.
  - Impact statistics:
    - Considering 2003 as start date, real GDP per capita in Germany grew by about 11 percent in 2003-2008, compared to 8.3 percent in 1997-2002.
    - Growth of average real GDP per capita in all advanced economies slowed from 10.6 percent in the pre-reform period to 8.7 percent in the post-reform period.
    - Synthetic control analysis: GDP per capita gap between Germany and its synthetic counterpart rose from 4.2 percent in 2003 to 9.3 percent in 2008.
    - Placebo experiment: 18 percent of placebo units showed a larger estimated impact of reform than Germany.
    - Placebo units (p-value): 22 (0.18).

### Methodological notes and statistical significance
- Approach: Synthetic control method used to construct counterfactuals, taking the starting year of the reform wave as the treatment date.
- Sample: Six individual cases analyzed—New Zealand, Australia, The Netherlands, Denmark, Ireland (1990s product and labor market reform packages) and Germany (Hartz reforms in the 2000s).
- Statistical significance:
  - In four out of six cases there is some evidence of a pick-up in GDP per capita relative to the counterfactual within five years after the reform start.
  - Exceptions: Denmark (no estimated effect relative to the counterfactual, although GDP per capita growth did pick up) and New Zealand (initial pick-up smaller than the no-reform counterfactual with some later catch-up).
  - With the exception of Ireland, placebo experiments do not permit claiming statistical significance of the results.

### Tentative policy conclusions
- Major reform packages are typically implemented gradually over several years and often take time to pay off, making it difficult to identify systematically large effects in their immediate aftermath.
- Supportive macroeconomic policies seem to enhance the short-term pay-off from reforms. New Zealand—where macroeconomic support was lacking when major reforms were implemented in the early 1990s—illustrates this point.

*International Monetary Fund Working Paper (excerpt).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16171.pdf_