## _sdn1514

## Source details

**Canonical URL:** [_sdn1514](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1514.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1514.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1514.pdf.json)

---

### Introduction: scope and measures
- Study period and sample:
  - 1980–2010.
  - 20 advanced economies: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States.
- Inequality measures considered:
  - Income share of the top 10 percent earners.
  - Gini coefficient (gross and net income).
- Institutional focus:
  - Labor market institutions (notably union density and the minimum wage) and their relation to the distribution of incomes.
- Motivating alternative explanations referenced:
  - Skill-biased technological change (SBTC), globalization, financial deregulation, and top marginal personal income tax rates.

### Main empirical findings
- Overall patterns:
  - The rise of inequality in the sample has been driven mainly by the upper part of the income distribution, particularly increases in income shares of top 10 percent earners.
  - Since the mid-1990s the Gini and top income share statistics have diverged; the Gini increased at a significantly slower pace than top income shares.
- Union density:
  - The decline in union density is strongly associated with the rise of top income shares.
  - Robust negative coefficients on union density across specifications for top 10 percent income share and Gini of gross income.
  - Benchmark magnitude: a 10 percentage point decline in union density is associated with a 5 percent increase in the top 10 percent income share.
  - Table 1 (Top 10 income share) Union density coefficient = -0.462 (t-statistic (8.62)**).
  - Table 2 (Gini of gross income) Union density coefficient = -0.308 (t-statistic (5.59)**).
  - Joint estimation for Gini of net income Union density coefficient = -0.29 (t-statistic (5.87)**).
- Minimum wage:
  - A 10 percentage point decline in the ratio of the minimum wage to the median wage is related to a 5 percent increase in the Gini coefficient of gross income.
  - Minimum wage strongly associated with the Gini of gross income but not with the top income share.
  - Benchmark (Ln(Gini of gross income)) Minimum wage coefficient = -0.365 (t-statistic (3.86)**).
- Other institutional and structural factors:
  - ICT (Ln(share ICT in K stock)) positively associated with top 10 income share: coefficient = 0.065 (t-statistic (6.34)**).
  - Financial reform (Ln(financial reform)) positively associated with top 10 income share: coefficient = 0.060 (t-statistic (3.74)**).
  - Top marginal personal income tax rate negatively associated with inequality: Top tax coefficient (Top 10 income share) = -0.115 (t-statistic (4.18)**).
  - Employment protection and unemployment benefits effects are mixed and not robust.

### Event analysis and distributional dynamics
- Event study (13 large declines in union density identified using Bry and Broschan algorithm):
  - Average decline in union density: 8 percentage points over a five-year period.
  - Top income shares rise by 1.8 percentage points five years after the event (average).
  - Labor income share declines by around 3 percentage points after five years.
  - Relative compensation in finance increases by about 13 percent after five years.
  - No clear break from pre-event trend for Gini coefficients in event averages.
  - Caveat: small number of events and no control for contemporaneous changes in other determinants in the event analysis.
- Decile analysis (distributional implications):
  - Higher union density associated with higher net income shares for deciles 1 through 7 and lower shares for deciles 8 through 10.
  - Example coefficients (selected years, 1981–2010):
    - Bottom Decile: 0.021 (t-statistic (7.59)**)
    - Decile 5: 0.016 (t-statistic (9.82)**)
    - Decile 9: -0.022 (t-statistic (8.00)**)
    - Top Decile: -0.085 (t-statistic (8.03)**)

### Quantitative magnitude and contributions (1980–2010)
- Contribution estimates:
  - On average, decline in union density explains about 40 percent of the 5 percentage point increase in the top 10 percent income share.
  - This contribution rises to over 50 percent when controlling for sectoral employment shifts.
  - Decline in union density explains about half of the increase in the Gini of net income.
- Minimum wage contributions:
  - In countries where the minimum wage declined the most, it accounts for about 2 percentage points of the increase in the Gini coefficient.
  - Where the minimum wage rose substantially, it reduced the Gini coefficient by about 2 percentage points.
- Counterfactual Gini calculation:
  - Under assumptions of a two-class economy and constant inequality within the bottom 90 percent since 1995, the evolution of the top 10 percent income share implies the Gini should have increased by at least 2.3 percentage points since 1995.
  - Note appended statements: "1.3 percentage points above the observed change." and "_sdn1514 - 1.3 percentage points above the observed change."

### Mechanisms and proposed channels
- Wage compression:
  - Unions and higher minimum wages compress the wage distribution; higher minimum wages reduce wage inequality.
- Unemployment effects:
  - Labor market institutions that compress wages can have ambiguous effects on unemployment; observed adverse unemployment effects of unions and minimum wages are generally modest within observed ranges.
- Redistribution and political influence:
  - Strong unions may induce more redistribution via political mobilization (Power Resources Theory (PRT)).
- Bargaining power and top income shares:
  - Declining union density may reduce bargaining power of average wage earners, lowering the labor share and increasing the capital income share; capital income is more concentrated, which can raise top income shares.
  - Weaker unions may reduce workers’ influence on corporate decisions, including top executive compensation.
  - Alternative interpretation: weaker unions may increase managerial freedom and productivity, potentially justifying higher top executive remuneration.

### Robustness and sensitivity
- Robustness checks where union density–top share relationship remains:
  - Controlling for globalization and SBTC.
  - Political orientation of governments and social preferences toward inequality.
  - Sectoral employment shares (industry and services).
  - Share of hours worked in finance: a one percentage point increase in finance hours share associated with a 7 percent increase in top 10 percent income share; union density results remain robust.
  - Higher education shares show no significant relation with top 10 percent income share.
  - Instrumental variables: results robust using lagged union density (first three lags) and a more sophisticated instrument set.
  - Alternative estimation methods: clustered standard errors, Driscoll-Kraay, random effects, between effects, cross-section differences, first differences with fixed effects.
- Robustness caveats:
  - Relation between union density and Gini of gross income is weaker and less robust than with top 10 percent income share.
  - Excess collective bargaining coverage results less stable and not included in benchmark.
  - Minimum wage coding issue: countries without a statutory national minimum wage coded as zero may understate minimum wage effects; restricting to countries with statutory minimum wage increases estimated magnitude.

### Policy considerations, trade-offs, and implications
- If interpreted causally:
  - Results suggest higher unionization and higher minimum wages can help reduce inequality.
- Trade-offs and country specificity:
  - Not a blanket recommendation—strong unions or high minimum wages can lead to high structural unemployment (especially for youth) and loss of competitiveness in some countries (for example, some southern European countries).
  - Minimum wages that are too high can lead to high unemployment among unskilled workers and reduced competitiveness.
  - Policy decisions on labor market institutions should be made on a country-by-country basis, taking into account macroeconomic stability, competitiveness, growth, and unemployment.
- Multi-pronged approach:
  - Combine labor market institution considerations with tax reform and curbing excesses associated with financial deregulation to address rising inequality.
- Policy tools highlighted:
  - Strengthening redistributive fiscal instruments (top marginal tax rates).
  - Supporting minimum wages where appropriate.
  - Addressing institutional changes in union representation and collective bargaining structures.
- Welfare and interpretation caveats:
  - Need to distinguish "good inequality" (productivity-based) from "bad inequality" (rent extraction).
  - Weaker unions may sometimes represent rent extraction benefiting some workers at the expense of others (for example, the young).

### Conclusion: synthesis
- The rise of inequality in most advanced economies was driven by the upper part of the income distribution, largely increases in income shares of the top 10 percent earners.
- The erosion of labor market institutions—particularly declining union density and, in many cases, declining minimum wages— is associated with increased inequality and weaker redistribution over 1980–2010.
- Other contributors include technological change (ICT share), globalization (China export share interaction), financial liberalization, and lower top marginal tax rates.
- Policy responses should be country-specific, balancing potential benefits for redistribution against risks to employment, competitiveness, and growth.

*Source: EXECUTIVE SUMMARY and CONCLUSION of "INEQUALITY AND LABOR MARKET INSTITUTIONS", International Monetary Fund (excerpts and empirical results, 1980–2011).*

### EXECUTIVE SUMMARY ______________________________________________________________________________ 4

### EXECUTIVE SUMMARY

### INTRODUCTION
- Study period and sample:
  - 1980–2010.
  - 20 advanced economies: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States.
- Inequality measures considered:
  - Income share of the top 10 percent earners.
  - Gini coefficient (gross and net income).
- Motivation:
  - Traditional explanations: skill-biased technological change (SBTC) and globalization.
  - Institutional explanations emphasized: financial deregulation and lower top marginal personal income tax rates.
  - This paper focuses on labor market institutions (notably union density and the minimum wage) and their relation to the distribution of incomes.

### MAIN FINDINGS
- Overall patterns:
  - The rise of inequality in the sample has been driven mainly by the upper part of the income distribution, particularly increases in income shares of top 10 percent earners.
  - Since the mid-1990s the Gini and top income share statistics have diverged; the Gini increased at a significantly slower pace than top income shares.
- Union density:
  - The decline in union density (the fraction of union members in the workforce) is strongly associated with the rise of top income shares.
  - Robustness checks were performed to mitigate concerns about omitted variables, endogeneity, and estimation method.
  - Evidence indicates unions can affect redistribution through their influence on public policy.
- Minimum wage:
  - Reductions in the minimum wage relative to the median wage are related to significant increases in inequality.
- Collective agreements coverage:
  - Some evidence that collective agreements coverage in excess of union density is associated with higher inequality, likely due to higher unemployment.
- Other labor market institutions:
  - Empirical evidence concerning the effects of unemployment benefits and employment protection on inequality and redistribution is not robust.
- Other institutional factors:
  - Financial deregulation and lower top marginal tax rates are confirmed to be related with higher inequality.

### PROPOSED CHANNELS / MECHANISMS
- Wage distribution:
  - Unions and higher minimum wages compress the wage distribution; higher minimum wages reduce wage inequality.
- Unemployment effects:
  - Labor market institutions that compress wages can also have ambiguous effects on unemployment; the evidence of adverse unemployment effects of unions and minimum wages is generally modest within observed ranges.
- Redistribution and political influence:
  - Strong unions may induce more redistribution via political mobilization (Power Resources Theory (PRT)).
- Bargaining power and top income shares:
  - Declining union density may reduce bargaining power of average wage earners, lowering the labor share and increasing the capital income share; capital income is more concentrated, which can raise top income shares.
  - Weaker unions may reduce workers’ influence on corporate decisions, including top executive compensation.
  - Alternative interpretation: weaker unions may increase managerial freedom and productivity, which could justify higher top executive remuneration.
- Empirical support note:
  - The strong negative relationship between unionization and top earners’ income shares is a novel and puzzling result that challenges the view that labor market institutions only affect low- and middle-wage workers.

### POLICY CONSIDERATIONS AND CAUTIONS
- Interpretation:
  - If interpreted causally, results suggest higher unionization and higher minimum wages can help reduce inequality.
- Trade-offs and country specificity:
  - Not a blanket recommendation—strong unions or high minimum wages can lead to high structural unemployment (especially for youth) and loss of competitiveness in some countries (for example, some southern European countries).
  - Minimum wages that are too high can lead to high unemployment among unskilled workers and reduced competitiveness.
  - Policy decisions on labor market institutions should be made on a country-by-country basis, taking into account macroeconomic stability, competitiveness, growth, and unemployment.
- Multi-pronged approach:
  - Results support combining labor market institution considerations with tax reform and curbing excesses associated with financial deregulation to address rising inequality.

### QUANTITATIVE ILLUSTRATION
- Counterfactual Gini calculation:
  - Given assumptions of a two-class economy (top 10 percent and bottom 90 percent) and constant inequality within the bottom 90 percent since 1995, the evolution of the top 10 percent income share implies the Gini should have increased by at least 2.3 percentage points since 1995.

*Source: EXECUTIVE SUMMARY of "INEQUALITY AND LABOR MARKET INSTITUTIONS", International Monetary Fund*

### 1.3 percentage points above the observed change.

### _sdn1514 - 1.3 percentage points above the observed change.

### Evolution of inequality in advanced economies (1980–2011)
- The increase in gross income inequality is on average larger than in net income inequality; cross-country dispersion of changes in net income inequality is smaller than for gross income measures.
- Net income inequality has increased in almost every year since 1980.
- Gross earnings differentials between the 9th and 5th deciles increased over four times as much as the differential between the 5th and 1st deciles (Figure 2).
- Net income shares from the Luxembourg Income Survey indicate that income shares of the top 10 percent have increased at the expense of all other income groups.

### Changes in labor market institutions (1980–2011)
- Union density declined steadily in most advanced economies, including after the mid-1990s.
- The minimum wage relative to the median wage evolved heterogeneously: declining in the United States and rising in France; some countries have no statutory national minimum wage (AUT, DNK, FIN, DEU, ITA, NOR, SWE, CHE).
- Clear trend decline in top marginal tax rates and liberalization of financial markets until the mid-1990s.

### Simple correlations and descriptive associations
- Strong negative relation between the top 10 percent income share and union density (within and across countries). Reported regression line: y = -0.005x + 3.633, R² = 0.4343 (All variation).
- Gini of gross income also negatively related to union density, but the relationship is somewhat weaker and mostly present within countries.
- Correlations for minimum wage and various inequality measures are mixed.

### Event analysis: large declines in union density
- Identification and sample:
  - Bry and Broschan algorithm used to identify turning points.
  - 13 events selected with average decline in union density of 8 percentage points over a five-year period.
- Key event findings (averages across 13 episodes):
  - Top income shares rise by 1.8 percentage points five years after the event.
  - No clear break from pre-event trend for Gini coefficients.
  - Labor income share declines by around 3 percentage points after five years.
  - Relative compensation in finance increases by about 13 percent after five years.
- Caveat: small number of events and no control for contemporaneous changes in other determinants in the event analysis.

### Panel regression analysis: specification and controls
- Simultaneous system estimated by three-stage least squares (3SLS) with country and year fixed effects.
- Sample and periods:
  - Top income share equations: sample of 18 countries, period 1981–2010.
  - Gini of gross and net income system: sample of 20 economies, period 1981–2010.
- Controls included:
  - Technology: share of information and communications technology (ICT) capital in total capital stock.
  - Globalization: share of China in world exports interacted with lagged income per capita.
  - Financial reform index (Abiad, Detragiache, and Tressel, 2008).
  - Top marginal personal income tax rate.
  - Banking crisis dummy.
  - Subset for redistribution equations: union density, unemployment benefits, top marginal tax rate.

### Baseline quantitative results (selected magnitudes)
- Union density:
  - A 10 percentage point decline in union density is associated with a 5 percent increase in the top 10 percent income share (benchmark estimate).
  - Union density coefficients are negative and statistically significant across specifications for top 10 percent income share and for the Gini of gross income.
  - In Table 1 (benchmark, Top 10 income share), coefficient on Union density = -0.462 (t-statistic (8.62)**).
  - In Table 2 (Gini of gross income), coefficient on Union density = -0.308 (t-statistic (5.59)**).
  - In joint estimation for Gini of net income, Union density coefficient = -0.29 (t-statistic (5.87)**).
- Minimum wage:
  - A 10 percentage point decline in the ratio of the minimum wage to the median wage is related to a 5 percent increase in the Gini coefficient of gross income.
  - Minimum wage strongly associated with the Gini of gross income but not with the top income share.
  - In Table 1 (Ln(Gini of gross income), benchmark), Minimum wage coefficient = -0.365 (t-statistic (3.86)**).
- Other determinants:
  - ICT share (Ln(share ICT in K stock)) is positively associated with top income share and Gini; e.g., in Table 1 Ln(share ICT in K stock) = 0.065 (t-statistic (6.34)**) for top 10 income share.
  - Financial reform (Ln(financial reform)) positively associated with inequality; e.g., Table 1 Ln(financial reform) = 0.060 (t-statistic (3.74)**) for top 10 income share.
  - Top tax rates exhibit redistributive association: Top tax coefficient negative and significant across specifications (e.g., Table 1 Top tax = -0.115, (4.18)** for top 10 income share).
  - Employment protection and unemployment benefits showed mixed and less robust effects.

### Redistribution and union density
- Net income inequality is strongly associated with gross income inequality; the coefficient less than one reflecting redistribution.
- For a given Gini of gross income, lower union density is correlated with a higher Gini of net income, supporting the hypothesis that unions influence redistribution.
- Excess collective bargaining coverage (coverage minus union density) can be positively associated with higher inequality in some specifications, possibly via higher unemployment, but this result is not robust across all tests.

### Robustness checks and additional tests
- Results for union density and top 10 percent income share are robust to:
  - Controlling for globalization and SBTC (skill-biased technological change).
  - Political orientation of governments and social preferences toward inequality.
  - Sectoral employment shares (industry and services).
  - Share of hours worked in finance: one percentage point increase in finance hours share associated with a 7 percent increase in top 10 percent income share; union density results remain robust.
  - Higher education shares: no significant relation with top 10 percent income share; union density result unaffected.
  - Instrumental variables for union density: results robust using lagged union density (first three lags) and a more sophisticated instrument set.
  - Alternative estimation methods (clustered standard errors, Driscoll-Kraay, random effects, between effects, cross-section differences, first differences with fixed effects).
- Robustness caveats:
  - Relation between union density and Gini of gross income is weaker and less robust than with top 10 percent income share.
  - Excess collective bargaining coverage results less stable and not included in benchmark.
  - Minimum wage data coding: countries without a statutory national minimum wage coded as zero may cause underestimation of minimum wage effects; restricting to countries with statutory minimum wage increases estimated magnitude.

### Distributional implications (decile analysis)
- Higher union density is associated with higher net income shares for deciles 1 through 7 and lower shares for deciles 8 through 10.
- Coefficient on union density for the 10th decile is much larger (in absolute value) than for any other decile, supporting the baseline result that union density decline is linked to rising top income shares.
- Table 4 (selected years, 1981–2010) reports union density impacts across deciles; example coefficients for union density on deciles:
  - Bottom Decile: 0.021 (t-statistic (7.59)**)
  - Decile 5: 0.016 (t-statistic (9.82)**)
  - Decile 9: -0.022 (t-statistic (8.00)**)
  - Top Decile: -0.085 (t-statistic (8.03)**)

### Magnitude and contribution analysis (1980–2010)
- On average, decline in union density explains about 40 percent of the 5 percentage point increase in the top 10 percent income share (top panel Figure 7).
  - This contribution rises to over 50 percent when controlling for sectoral employment shifts.
- Decline in union density explains about half of the increase in the Gini of net income.
- Contributions of minimum wage changes to the Gini of gross income vary by country:
  - In countries where the minimum wage declined the most, it accounts for about 2 percentage points of the increase in the Gini coefficient.
  - Where the minimum wage rose substantially, it reduced the Gini coefficient by about 2 percentage points.
- Figure 7 aggregates contributions as the product of observed change in each variable over 1980–2010 and its estimated coefficient, averaged across advanced economies.

### Policy-relevant implications and interpretation
- Declining union density is a key labor market institutional change associated with rising top income shares and with weaker redistribution, making it an important explanatory factor for the rise in inequality in advanced economies over 1980–2010.
- Minimum wage dynamics matter for overall inequality (Gini of gross income), though their aggregate contribution is heterogeneous across countries.
- Other factors—technological change (ICT share), globalization (China export share interaction), financial liberalization, and top marginal tax rates—also significantly contributed to the rise in inequality.
- Collective bargaining coverage that substantially exceeds union density (excess coverage) can be associated with higher inequality, potentially via higher unemployment; however, evidence is less robust.
- Policy tools that can mitigate rising inequality include strengthening redistributive fiscal instruments (top marginal tax rates), supporting minimum wages where appropriate, and addressing institutional changes in union representation and collective bargaining structures.

*Source: IMF — Inequality and Labor Market Institutions (excerpts and empirical results, 1980–2011). *

### CONCLUSION

### CONCLUSION

### Main findings
- The rise of inequality in most advanced economies was driven by the upper part of the income distribution, owing largely to the increase in income shares of the top 10 percent earners.
- There is strong evidence that the erosion of labor market institutions in the advanced economies examined is associated with an increase of income inequality.
- If interpreted as causal, results suggest:
  - The weakening of unions contributed to the rise of top earners’ income shares and less redistribution.
  - Eroding minimum wages increased overall inequality considerably.
- Some evidence indicates that collective agreements coverage in excess of union density may be related to higher inequality, likely through higher unemployment.
- Other institutional factors confirmed to be related to higher inequality include financial deregulation and lower top marginal tax rates.

### Interpretation and possible channels
- The strong link found between unions and top income shares contrasts with much of the literature, which has not extensively explored the relationship between labor market institutions and top income shares.
- Mechanisms discussed:
  - De-unionization may restrain earnings at the middle and bottom of the distribution, causing income shares of top earners to rise.
  - Weaker unions may increase the share of capital income, which tends to be more concentrated than labor income.
  - Lower union density may reduce workers’ influence on corporate decisions, including those related to top executive compensation.
- These hypotheses have been discussed but not extensively tested empirically; future work is planned to identify the most relevant channels.

### Welfare implications and policy trade-offs
- Whether policy action is warranted depends on the welfare implications of the rise in inequality attributable to weakened labor market institutions; assessment must be country by country.
- Distinctions to consider:
  - The rise in top earners’ income shares could reflect a relative increase in their productivity ("good inequality").
  - Top earners’ compensation may exceed their contribution to output, reflecting rent extraction ("bad inequality").
  - Inequality could allow top earners to manipulate economic and political systems, yielding grounds for policy action.
- Results suggest that higher unionization and minimum wages can help reduce inequality.
- Counterpoint: where unions primarily represent some workers’ interests, this can function as rent extraction at the expense of other workers (for example, the young).

### Policy recommendations and cautions
- A country-by-country approach is needed, weighing potential conflicts between strengthening labor market institutions and other macroeconomic objectives.
- Particular caution where strong unions and high minimum wages have been associated with high structural unemployment and losses of competitiveness (for example, in some southern European countries).
- Addressing the increase in inequality may also require:
  - Tax reform.
  - Regulations to curb excesses in the financial sector.

*Source: CONCLUSION, _sdn1514 - CONCLUSION*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1514.pdf_