## _wp14184

## Source details

**Canonical URL:** [_wp14184](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14184.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14184.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14184.pdf.json)

---

### 2.1 Evolution of China’s Minimum Wage Policy — Background and Institutional Evolution
- China approved the International Labor Organization (ILO)’s Minimum Wage-Fixing Machinery Convention (1928) in early 1984.
- July 1994 Labor Law: requirement to implement a system of guaranteed minimum wages; Article 48 required firms to comply with local minimum wage regulations.
- Subnational authority:
  - Provincial governments authorized to set their own minimum wage standards; cities and counties can negotiate local minimum wages with higher authorities.
  - As of 2007 administrative divisions: 2,867 counties, 333 cities, and 31 provinces.
- Key reform milestones:
  - 1994: Minimum-Wage Law enacted.
  - March 2004: Ministry of Labor directive establishing more comprehensive minimum wage coverage and increased non-compliance penalties.
  - 2008: Labor Contract Law restructured and emphasized the importance of minimum wage policies; Ministry of Human Resources and Social Security provided guidelines allowing delays in adjustments during the global financial crisis.
- Enforcement and frequency:
  - Before 2003 adjustments were less frequent; in 1998 only one fifth of all counties adjusted their minimum wages.
  - By the end of 2007 all provinces had implemented the new minimum wage regime and enhanced enforcement measures.

### 2.1 Major Elements of the 2004 Reform (Ministry of Labor Directive)
- Extension of coverage to town-village enterprises and self-employed business.
- Creation of a new standard for hourly minimum wages.
- Penalties for violators increased from 20%-100% to 100%-500% of the wage owed.
- Minimum-wage adjustment required at least once every two years.
- Post-2004 adjustment process:
  - Provincial government drafts proposal → local labor unions, trade and business communities discuss → Ministry of Labor reviews and provides revisions/comments → if no revision requests within 28 days, provincial government authorized to adjust and publish in local newspapers within 7 days.

### Data Specifics and Measurement Conventions
- Data source and span:
  - Minimum wage data collected by the Ministry of Human Resources and Social Security from official reports of local governments; data span from 1992 to 2012 and contain detailed information on adjustments at the county level.
- Measurement conventions:
  - Minimum wages specified as monthly wages, part-time hourly wages, and full-time hourly wages; monthly minimum wages used in estimation.
  - Most counties set full-time hourly minimum wages based on monthly minimum wages divided by a factor of around 175.
  - Implementation dates known; monthly aggregated minimum wages converted to annual minimum wages for matching with annual firm data. Annual minimum wages reflect effective total minimum wage for each year.
- Geographic resolution:
  - County-level data reduce geographic unit by a factor of about ten compared with municipal cities, increasing variation and improving identification through county-wise comparisons.

### Firm and Sample Coverage (Manufacturing Sector)
- Firm data source: Annual Survey of Industrial Firms (ASIF), also known as Chinese Industrial Enterprise Database (CIED).
- Sample period and size:
  - Sample spans from 1998 to 2007.
  - After panel construction the dataset contains 2,043,435 observations in the unbalanced panel for ten years.
- Coverage and representativeness:
  - The 2004 survey covers more than 91 percent of China’s manufacturing output and 71 percent of manufacturing employment.
  - As reported by the 2004 economic census, total employment of China’s manufacturing sector amounts to 93.4 million.
- Attrition and sampling:
  - Attrition rate in the sample varies from 8.2 percent to 20.3 percent over the nine years of 1999-2007.
  - Rate of re-entry is around 10 percent over time.
- Firm-level measures:
  - ASIF provides end-of-year balance sheets, input and output, average number of employees over a year, and annual wage bills.
  - Average wage per employee used as an indicator of firm exposure to minimum wage shocks (number of employees paid below minimum wage not identified).

### Empirical Identification and Geographic Design
- Matching strategy:
  - County minimum wages matched with a novel dataset of neighbor county-pairs (each county paired with adjacent counties) to focus on within-neighbor-pair variation.
- Neighbor definition:
  - Neighbors share a border on land, rivers, or coastal waters and have centroid distance no more than eighty kilometers.
  - Median number of county neighbors in the data is six.
- Rationale:
  - Geographic proximity used to control for local labor market conditions and reduce differences in market conditions between treatment and control regions.
  - Adjacent counties considered better comparison groups than adjacent cities.
  - 22 percent of cities have uniform minimum wages across counties and sub-districts; 78 percent show within-city variation.

### Main Empirical Findings (Overview and Interpretation)
- Elasticity of minimum wages on firm employment (main sample): point estimate −0.103 (statistically significant).
- Interpretation:
  - Employment adjustment is a channel through which firms accommodated labor cost increases and stricter regulation.
  - Negative elasticity implies a large economic impact given sample coverage.
- Enforcement intensity:
  - Rise in enforcement intensity from 2004 onwards amplified the negative effect after about one year.
  - Strengthened enforcement led to reductions in employment by raising wages of low-income employees.
- Heterogeneity:
  - Firms with higher wage levels and higher profit margins exhibit less negative elasticities to minimum wage increases.
- Context on rising wages:
  - From 1998 to 2010 the average growth rate of real wages was 13.8 percent, exceeding real GDP growth and labor productivity growth.

---

### 3.3 Macroeconomic Factors — Theoretical Background and City/Provincial Variables
- Data sources and coverage:
  - Major source: China Statistical Yearbook Database from CNKI.
  - City panel: 337 cities over 1990-2012; includes four directly controlled municipalities and 333 prefectural cities.
- Wages context:
  - Ratio between manufacturing firm median wages and city average wages was 65 percent in 2004 and remained stable over 2001-2007.
  - In 2004, manufacturing firm wages were below city average wages in 96 percent of cities.
- Limitations:
  - No complete panel of economic variables at county level; some variables like consumer price index only at province level.
- Theoretical labor supply cases:
  - Perfectly elastic labor supply (competitive model): increase in market wage reduces firm employment (cost channel).
  - Imperfectly elastic labor supply (monopsony): minimum wage increase may reduce firm marginal cost and could raise labor demand.
- Labor demand production function and implications:
  - Production function: Y = A K^α L^β M^(1−α−β)
  - Labor demand expression presented; fundamentals include firm productivity A, price elasticity σ, and market conditions.
  - Competitive labor market implies minimum wages negatively correlated with firm employment; elasticity of labor demand to wage rate stated as −(1+β(σ−1)) (as presented).
- Empirical consequences:
  - Linear regression model to estimate wage elasticity of firm employment; include market prices and aggregate demand; control for lagged firm employment for labor adjustment frictions.

### Determinants of Local Minimum Wages (Empirical Implementation)
- Policy process captured by variables: MW = f(pC, S, A, U, E, a)
  - C: average level of consumption (city average wages; provincial CPI as deflator)
  - S: social security
  - A: local average wage
  - U: unemployment rate
  - E: general condition of local economy
  - a: other factors
- Four categories of empirical variables:
  1. Local labor income and living costs: city average wages per employee; provincial CPI.
  2. Local economic growth prospects: lagged GDP per capita growth and fixed asset investment.
  3. Industrial policy and sector composition: output shares of secondary and tertiary industries; growth rate of FDI.
  4. Local labor market conditions: lagged labor force growth and unemployment rate.
- City-level regression specification:
  - ln MW_ct = α + X_{i,t−1} β + μ_c + τ_t + ε_ct
  - City sample: minimum wage data covers 2,374 county-level districts in all the cities; regressions use 346 cities; sample period 1994–2011.
- Key empirical findings:
  - Median minimum wages around one third of median city wages per employee (1994–2011).
  - Median city wages per employee one quarter higher than median city GDP per capita.
  - After 2004 (post-enforcement reform), coefficients on city wages, GDP growth, and fixed asset investment statistically significant at 1%.
  - After 2004, living costs (city average wages, log) elasticity with respect to minimum wages = 2.96 (Column 1).
  - Fixed-asset investment positively related to minimum wages; lagged GDP per capita growth showed negative effect (Column 2).
  - Growth rate of labor force and unemployment rate not statistically significant (Column 3).
  - Within-city variation of minimum wages well explained; cross-city variation less so.
  - Conclusion: controlling for main city-level variables makes remaining changes in minimum wages more plausibly exogenous for firm-level regressions.

---

### 5.2 Neighbor County Pairs — Strategy, Rationale, and Empirical Model
- Strategy:
  - Use neighbor county-pairs akin to a regression discontinuity design to control for unobservable factors that change over time.
  - Treatment threshold selected as thirty percent.
  - Many pairs cross provincial borders; significant differences observed in provinces such as Liaoning, Hunan, Jiangxi, and Guangdong.
  - Repeated counties assigned lower weights; standard errors adjusted.
- Rationale and advantages:
  - Within county-pair variation controls for all trends experienced by both counties in a pair.
  - With one neighbor-pair in sample, approach reduces to difference-in-difference.
  - Controlling for neighbor county-pair fixed effects addresses positive correlation bias between unobserved regional determinants of firm employment and local minimum wage setting; coefficients of minimum wages are more negative when pair fixed effects included.
- Baseline empirical model (first-difference with Difference GMM instruments):
  - lnL_it = α + β_1 lnL_i,t-1 + X_it β + X_ct β + γ lnMW_ct + μ_i + μ_p + τ_t + ε_it
  - γ measures elasticity of minimum wages on firm employment.
  - Firms that changed locations over the sample period were dropped.
- Key variables:
  - Dependent variable: lnL_it (log of firm employment).
  - Firm wage W_i: wage bill divided by employment (generally not used as explanatory due to joint determination).
  - Labor cost measured by county minimum wages (MW), city average wages, industry average wages.
  - Other controls: province fixed investment price index, industry intermediate input price index, industry output, city GDP per capita, HHI (price elasticity proxy), labor income share, profit margin, ownership dummies, export-to-sales ratio, lagged firm employment L_i,t-1 (instrumented by L_i,t-2), firm size S_i,t-1.

---

### 6 Main Results — Average Effects, Timing, and Robustness
- Analysis period: 2000-2007.
- Estimation steps:
  1. Average effects of minimum wage on firm employment and firm wages (2000-2007).
  2. Heterogeneous effects by firm characteristics (particularly firm wages).
  3. Robustness and diagnostics.
- Average elasticities and timing:
  - Minimum wage elasticity for 2000-2004: −0.027 (not statistically significant).
  - Minimum wage elasticity for 2005-2007: −0.067 (statistically significant at 1%).
  - Subsample estimate for 2005-2007 (alternative specification): −0.103. Interpretation: a 10 percent minimum wage hike → 1.03 percent reduction in hiring.
  - Subsample estimate for 2000-2004 (alternative specification): 0.022 (insignificant).
  - Whole sample without pair dummies: coefficient = −0.031 (significant at 5%).
  - With pair dummies: coefficient = −0.103 (significant at 1%).
- Time pattern:
  - Effects insignificantly negative 2000-2002; turn more negative in 2003; highly significant and more negative after 2005, suggesting a lagged effect of the March 1, 2004 reform.
  - Difference between elasticity 2000-2004 and 2005-2007 statistically significant.
- Robustness to regional controls:
  - For 2005-2007, coefficients vary from −0.059 to −0.103 across specifications, all statistically significant.
  - For 2000-2004, effects remain statistically insignificant across different macro controls.
- Firm-level covariates:
  - Lagged employment strongly predictive (slow employment adjustment).
  - Firm size (sales) negatively related to employment (coefficient smaller in magnitude than minimum wage effect).
  - Ownership:
    - State firms similar to foreign firms after tightening, but hired more before 2005.
    - Private firms hire fewer workers compared with foreign firms over 2000-2007.
    - HMT firms hire more only after 2005.
  - Profitability: higher profit margins associated with hiring more.
  - Exports: firms with more exports may hire fewer workers.
  - HHI: negative coefficient (in line with theory).
  - Industry labor share: positive coefficient.
- Interpretation of neighbor-pair fixed effects:
  - Including neighbor-pair fixed effects yields more negative elasticity (−0.103) than without (−0.031), consistent with controlling positive correlation bias.

### Effect of Minimum Wages on Firm Per-Employee Wages (Channel Evidence)
- Estimated wage regressions show:
  - Minimum wage elasticity on firm wages for 2005-2007: 0.349. Interpretation: minimum hikes explain one third of firm wage increase in this period.
  - City average wage elasticity: 0.491.
  - Industry wage elasticity: 0.132 (2005-2007); 0.013 (2000-2004).
- Firm characteristics on wages:
  - Large firms and private firms tend to pay lower wages.
  - Higher profitability associated with higher wage rates.
- Interpretation:
  - Minimum wage hikes substantially raise average firm wages, supporting the cost channel to employment.
  - Minimum wage effect on firm wages similar before and after enforcement tightening, but employment impact differs—possible roles for fringe benefits, price pass-through, and changing product market competitiveness.

---

### 6.2 Heterogeneous Effects — By Firm Wage Deciles and Other Characteristics
- Grouping by firm wage deciles (relative to city):
  - Period 2005-2007:
    - Top wage-decile firms: employment elasticity = −0.048 (insignificant).
    - Bottom wage-decile firms: employment elasticity = −0.153 (highly significant).
    - Interpretation: low-wage firms face binding labor demand constraints; high-wage firms show smaller or positive employment responses.
  - Period 2000-2004:
    - Top decile firms: elasticity = 0.096 (significant).
    - Bottom decile firms: elasticity = −0.034 (insignificant).
    - Monotone relationship between minimum wage elasticities and firm wage persists; average effect small pre-enforcement tightening.
- Full heterogeneous interactions (2005-2007):
  - MW × firm wage t-1: 0.117*** and 0.124*** in different specifications.
  - MW × sales t-1: −0.025***, −0.029***.
  - MW × SOE: −0.087***, −0.123***.
  - MW × profit margin t-1: 0.183***, 0.115***.
  - Interpretations:
    - State-owned enterprises reduce employment more than other firms in response to minimum wage hikes.
    - Lower-profit firms cut employment more than higher-profit firms.
    - No tendency for firms in concentrated industries to reduce hiring following minimum wage increases.
- Robustness checks:
  - Treatment dummy specification:
    - Treat t (binary) coefficient = −0.006.
    - Median of minimum wage hikes for treatment counties = 8 percent.
  - Sample attrition and Heckman correction:
    - Annual attrition ≈ 10 percent; re-entry ≈ 10 percent.
    - Probit predictors: larger sales, larger employment, higher profit margins → more likely to stay.
    - County minimum wages generally not statistically significant predictors of attrition except in 2003.
    - Heckman correction barely affects MW estimates; MWt remains −0.103*** after correction.
  - Placebo tests:
    - Real treatment: coefficient = −0.006.
    - Pseudo backward treatment (one year earlier): coefficient = 0.003 (statistically significant).
    - Interpretation: minimum wage adjustments are hard to predict one year before; placebo evidence supports causal interpretation of the real effect.

### Conclusion — Synthesis and Policy-Relevant Interpretations
- Historical and empirical summary:
  - Minimum wage legislation enacted in 1994; enforcement strengthened after 2004.
  - Using neighbor county pairs, average effects on firm employment:
    - Elasticity 2000-2004 = 0.022 (not statistically significant).
    - Elasticity 2005-2007 = −0.103 (statistically significant at 1%).
  - Minimum wage hikes increase profit margins and employee wages in treatment counties; firms are not uniformly worse off when facing local minimum wage increases.
- Policy-relevant interpretations:
  - Enforcement tightening increased the employment costliness of minimum wage policy, but lack of deterioration in firm profitability suggests government may have accommodated regulation to protect local firms.
  - Unclear whether such regulatory style is welfare-enhancing or distortionary.
  - For employees: incumbent workers benefit from minimum wage increases; workers who quit due to policy change may be adversely affected short-run; long-run effects on displaced workers not evaluated due to unobserved job reallocation.
  - Findings shed light on subsequent labor market regulations in China, such as the labor-contract law of 2008; labor market regulations can be binding but negative effects on firms can be mitigated by enforcement and government accommodation.

### Selected Key Statistics and Figures (Preserved Values)
- Trend of mean city annual (effective) minimum wages and mean city nominal wages over 2000-2011 (values in RMB): 252, 270, 288, 305, 334, 379, 420, 485, 558, 575, 654, 790, 676, 787, 897, 1,000, 1,144, 1,290, 1,481, 1,779, 2,076, 2,309, 2,615, 2,771.
- Table I: Percentage of counties with minimum wage hikes greater than thresholds (sample size 2,374) — selected years:
  - 1996: 36 (ą0 %), 25 (ą10%), 13 (ą20%)
  - 2004: 75, 42, 19
  - 2005: 85, 59, 38
  - 2006: 99, 46, 18
  - 2007: 99, 65, 35
  - 2011: 100, 97, 56
- Table II: Distribution of neighbor county differences in minimum wage hikes (2000-2007), where a hike is neighbor difference > 1%:
  - ą0% - 10%: 58%
  - 10% - 15%: 14%
  - 15% - 75%: 28%
- Key summary statistics (Table III, selected medians and means):
  - Monthly minimum wage: Median 330; Mean 393; STD 205; Min 125; Max 1,266
  - Monthly wage per employee: Median 934; Mean 1,227; STD 862; Min 193; Max 6,319
  - Monthly GDP per capita: Median 701; Mean 1,189; STD 1,359; Min 71; Max 5,361
  - Employees (firm-level): Median 120; Mean 285; STD 1,014; Min 2; Max 8,151
  - Monthly firm employee wage: Median 967; Mean 1,205; STD 981; Min 837; Max 3,693
  - Firm shares over years (1999, 2001, 2003, 2005, 2007): State firms 31%, 20%, 12%, 7%, 5%; Private firms 50%, 59%, 65%, 71%, 72%; HMT firms 11%, 12%, 12%, 11%, 11%; Foreign firms 8%, 9%, 10%, 12%, 12%

### Selected Econometric Estimates (Preserved Coefficients and Significance)
- Employment elasticities:
  - MW elasticity 2000-2004 (main): −0.027 (not significant).
  - MW elasticity 2005-2007 (main): −0.067***.
  - Subsample 2005-2007 (alternative): −0.103***.
  - Without pair dummies: MW = −0.031**; with pair dummies: MW = −0.103***.
- Year-specific MW coefficients (whole sample): 2000: −0.014; 2001: −0.002; 2002: −0.025; 2003: 0.048**; 2004: −0.045**; 2005: −0.083***; 2006: −0.058***; 2007: −0.086***.
- Firm wage response (2005-2007): MW elasticity on firm wages = 0.349***.
- Heterogeneous effects by wage deciles (2005-2007, selected):
  - Bottom decile (0-10%): −0.153***.
  - Top decile (90-100%): −0.048 (insignificant).
- Interaction estimates (2005-2007, Table X):
  - MW × firm wage t-1: 0.117***, 0.124***.
  - MW × sales t-1: −0.025***, −0.029***.
  - MW × SOE: −0.087***, −0.123***.
  - MW × profit margin t-1: 0.183***, 0.115***.

*Italic: Source content: _wp14184 (selected sections 2.1, 3.3, 5.2, 6, 6.2, References) from the provided PDF content unit.*

### 2.1    Evolution of China’s Minimum Wage Policy  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .4

### 2.1    Evolution of China’s Minimum Wage Policy

### Background and institutional evolution
- China approved the International Labor Organization (ILO)’s Minimum Wage-Fixing Machinery Convention (1928) in early 1984.
- In July 1994 the Labor Law stated a requirement to implement a system of guaranteed minimum wages; Article 48 required firms to comply with local minimum wage regulations.
- Provincial governments are authorized to set their own minimum wage standards; cities and counties can negotiate local minimum wages with higher authorities.
- Before 2003 adjustments of minimum wages were less frequent; in 1998 only one fifth of all counties adjusted their minimum wages.
- Key reform milestones:
  - 1994: Minimum-Wage Law enacted.
  - March 2004: Ministry of Labor issued a directive establishing more comprehensive minimum wage coverage and increased non-compliance penalties.
  - 2008: Labor Contract Law restructured and emphasized the importance of minimum wage policies; Ministry of Human Resources and Social Security provided guidelines allowing delays in adjustments during the global financial crisis.
- As of 2007 administrative divisions: 2,867 counties, 333 cities, and 31 provinces.

### Major elements of the 2004 reform (directive by the Ministry of Labor)
- Extension of coverage to town-village enterprises and self-employed business.
- Creation of a new standard for hourly minimum wages.
- Increase in penalty for violators from 20%-100% to 100%-500% of the wage owed.
- More frequent minimum-wage adjustment: once at least every two years.
- Post-2004 process: provincial government drafts a proposal → discussion by local labor unions, trade and business communities → Ministry of Labor reviews and provides revisions/comments → if no revision requests within 28 days, provincial government authorized to adjust and publish in local newspapers within 7 days.
- By the end of 2007 all provinces had implemented the new minimum wage regime and enhanced enforcement measures.

### Data specifics and measurement conventions
- Minimum wage data collected by the Ministry of Human Resources and Social Security from official reports of local governments; data span from 1992 to 2012 and contain detailed information on adjustments at the county level.
- Minimum wages specified as monthly wages, part-time hourly wages, and full-time hourly wages; the paper uses monthly minimum wages in estimation.
- Most counties set full-time hourly minimum wages based on monthly minimum wages divided by a factor of around 175.
- Because implementation dates of adjustments are known, monthly aggregated minimum wages are converted to annual minimum wages for matching with annual firm data; these annual minimum wages reflect the effective total minimum wage for each year.
- County-level data advantage: geographic unit reduced by a factor of about ten compared with municipal cities, increasing variation and improving identification through county-wise comparisons.

### Firm and sample coverage (manufacturing sector)
- Firm data source: Annual Survey of Industrial Firms (ASIF), also known as Chinese Industrial Enterprise Database (CIED).
- Sample spans from 1998 to 2007.
- Panel contains exactly the same number of observations used by NBS during all these years; after panel construction the dataset contains 2,043,435 observations in the unbalanced panel for ten years.
- Coverage and representativeness:
  - The 2004 survey covers more than 91 percent of China’s manufacturing output and 71 percent of manufacturing employment.
  - As reported by the 2004 economic census, total employment of China’s manufacturing sector amounts to 93.4 million.
- Attrition and sampling notes:
  - Attrition rate in the sample varies from 8.2 percent to 20.3 percent over the nine years of 1999-2007.
  - Rate of re-entry is around 10 percent over time, suggesting some firms left the sample due to sampling omissions rather than closure.
- Firm-level measures:
  - ASIF provides end-of-year balance sheets, input and output, average number of employees over a year, and annual wage bills.
  - Average wage per employee is used as an indicator of a firm’s exposure to minimum wage shocks, though the number of employees paid below minimum wage is not identified.

### Empirical identification and geographic design
- County minimum wages are matched with a novel data set of neighbor county-pairs (each county paired with adjacent counties) to focus on within-neighbor-pair variation.
- Geographic proximity is used to control for local labor market conditions and reduce differences in market conditions between treatment and control regions; adjacent counties are considered better comparison groups than adjacent cities for controlling time-varying unobservables.
- The county division, relative to prefectural or provincial divisions, helps exploit richer within-province variation in minimum wages; 22 percent of cities have uniform minimum wages across counties and sub-districts, while 78 percent show within-city variation.

### Main empirical findings and interpretation
- The elasticity of minimum wages on firm employment (main sample) is statistically significant with a point estimate of -0.103.
- Interpretation:
  - Employment adjustment is one channel through which firms accommodated labor cost increases and stricter regulation.
  - Given the sample includes almost all large manufacturing firms and a vast labor force, the negative elasticity implies a large economic impact.
- Enforcement intensity:
  - The rise in enforcement intensity from 2004 onwards amplified the negative effect after about one year.
  - Strengthened enforcement led to reductions in employment, operating through raising wages of low-income employees.
- Heterogeneity:
  - Firms with higher wage levels and higher profit margins exhibit less negative elasticities to minimum wage increases.
- Relation to other findings in China and the literature:
  - Results align with Wang and Gunderson (2011): negative employment effects in regions with slower growth and larger negative effects for non-state enterprises.
  - Fang and Lin (2013): minimum wage increases reduced employment for females, young adults, and less-skilled workers.
  - The paper’s findings are consistent with theoretical predictions from models of competition, (dynamic) monopsony, efficiency wage, and labor market search with frictions; cited theoretical work includes Manning (1995); Rebitzer and Taylor (1995); Bhaskar et al. (2002); Lang and Kahn (1998); Burdett and Mortensen (1998); Acemoglu (2001); Flinn (2006).
- Context on rising wages: from 1998 to 2010 the average growth rate of real wages was 13.8 percent, exceeding real GDP growth and labor productivity growth.

*Source: _wp14184 - 2.1    Evolution of China’s Minimum Wage Policy*

### 3.3    Macroeconomic Factors

### 3.3    Macroeconomic Factors

### Economic variables at the city and provincial level
- Major source: China Statistical Yearbook Database from CNKI.
- City database panel:
  - 337 cities over the period of 1990-2012, with interpolation for some missing values in early years.
  - These 337 cities include four directly controlled municipalities and 333 prefectural cities.
- Annual dataset can completely match up with the sample of manufacturing firms.
- Notes on wages:
  - The ratio between manufacturing firm median wages and city average wages was 65 percent in 2004 and remained stable over the period of 2001-2007.
  - In 2004, the wages of manufacturing firms were below city average wages in 96 percent of cities.
- Limitations:
  - Other than minimum wages, there is not a complete panel of economic variables at the county level.
  - Some variables, such as consumer price index, can only be observed at the province level.

### Geography and neighbor counties
- Neighbor county-pairs data set for the year 2007.
- Definition of neighbors:
  - Share same border either on land, in rivers, or in coastal waters, and
  - Distance between their centroids is no more than eighty kilometers.
- Median number of county neighbors in the data is six.
- Rationale: Geographic adjacency increases the likelihood that neighbor counties experience similar growth factors, especially unobservable shocks.

### Theoretical background
- Purpose: Lay micro-foundations of firm behavior to guide variable selection and sign predictions in regressions.
- Labor supply cases considered:
  - Perfectly elastic labor supply:
    - Marginal cost of labor equals market wage.
    - Firm hiring depends on productivity and product demand.
    - Cost channel implies an increase in the market wage must reduce a firm’s employment.
    - Corresponds to the competitive model of minimum wages.
  - Imperfectly elastic labor supply:
    - Firm may need to react to supply-side changes.
    - A minimum wage increase may reduce a firm’s marginal cost (though total labor cost rises), potentially raising labor demand and expanding hiring.
    - Corresponds to the monopsony model of minimum wages.
- References to literature: Brown (1999); Neumark and Wascher (2008); Schmitt (2013).

### Labor demand (theoretical specification)
- Production function assumed:
  - Y = A K^α L^β M^(1−α−β)
  - where Y, K, M, and L denote output, capital input, intermediate input, and labor input respectively; A is firm-specific productivity.
- Labor demand expression (as presented):
  - L = Ȳ^{1} A^{−1} (w/ P̄)^{−β/(1−σ)} (r/ P̄)^{−α/(1−σ)} (p_M/ P̄)^{−(1−α−β)/(1−σ)} [ (1−σ)/β (w/ P̄) ]^{σ/(σ−1)}  (equation displayed in source)
  - Fundamental determinants of firm employment: firm characteristics A, price elasticity σ, and market conditions Ȳ, w/ P̄, r/ P̄, and p_M/ P̄.
- Implications:
  - Competitive labor market implies minimum wages are negatively correlated with firm employment.
  - Elasticity of labor demand to the wage rate is −(1+β(σ−1)) (stated in text as ́p1`βpσ ́1qq).
- Empirical consequences:
  - Leads to a linear regression model to estimate wage elasticity of firm employment.
  - Need to include market prices and aggregate demand; other market factors (e.g., labor flow and growth) cause endogeneity in regional employment regressions.
  - Control for lagged firm employment to address labor adjustment frictions.

### Labor supply (theoretical considerations)
- Imperfectly elastic labor supply and monopsony:
  - A monopsonist may increase employment in response to a minimum wage hike by moving along the supply curve.
  - Card and Krueger (1995) provide a search model where elasticity of labor supply can be reasonably large, justifying positive or negligible correlations between minimum wages and employment.
- Empirical measures:
  - ASIF data lacks ideal indicators of labor supply shape.
  - Lagged firm wage used as an indicator of elasticity of labor supply if employment is supply-determined.
  - Control for market supply conditions such as labor force growth.
  - Use lagged firm wage to examine heterogeneous effects of minimum wages across firm groups.

### Identification and empirical framework

- Strategies to address endogeneity of minimum wages:
  - Control for economic determinants of minimum wages that firms can observe.
  - Use dynamic panel estimation: include lagged dependent variable so estimated elasticity can be interpreted as impact on employment growth.
  - Examine variation within neighbor county-pairs to exploit geographic proximity (method used since Card and Krueger (1994)).

- Policy process and adjustment of local minimum wages:
  - Minimum wage adjustment should consider variables: MW = f(pC, S, A, U, E, a)
    - C: average level of consumption
    - S: social security
    - A: local average wage
    - U: unemployment rate
    - E: general condition of local economy
    - a: other factors
  - Government trade-offs:
    - Freeze/slow adjustment to avert labor cost hikes and promote private investment (to avoid employment reductions).
    - Raise minimum wages due to social stability and citizen welfare concerns.
  - Empirical capture of considerations using four categories of variables:
    1. Local labor income and living costs:
       - City average wages per employee; provincial CPI as deflator.
       - CPI used as urban consumer price index; majority of manufacturing firms located in urban areas.
    2. Local economic growth prospects:
       - Lagged growth rate of GDP per capita and fixed asset investment.
    3. Industrial policy and sector composition:
       - Output shares of secondary and tertiary industries; growth rate of foreign direct investment (FDI).
    4. Local labor market conditions:
       - Lagged growth rate of the labor force and unemployment rate.
       - Ambiguous sign expected for labor force growth; unemployment expected to negatively affect willingness to raise minimum wages (measurement caveats noted).

- Empirical implementation: determinants of local minimum wages
  - City-level regression with fixed effects:
    - ln MW_ct = α + X_{i,t−1} β + μ_c + τ_t + ε_ct
    - MW_ct: minimum wage in city c at year t; μ_c: city fixed effect; τ_t: year fixed effects.
  - City sample and period:
    - Minimum wage data covers 2,374 county-level districts in all the cities.
    - City-level regressions use 346 cities; sample period from 1994 to 2011.
    - Columns 1–3 in reported regressions use sample 2004 to 2011 (post-enforcement reform); columns 4–6 use 1994 to 2003 (pre-enforcement tightening).
  - Key empirical findings:
    - Median minimum wages were around one third of the median of city wages per employee (over 1994-2011).
    - Median city wages per employee were one quarter higher than the median of city GDP per capita.
    - Growth of GDP per capita and FDI peaked in the run-up to the global financial crisis in 2008.
    - Share of tertiary industry stable; share of secondary industry rising since 2001.
    - Before enforcement tightening (pre-2004), coefficients on determinants are not statistically significant despite high within R^2, reflecting low frequency of adjustment.
    - After 2004, coefficients on city wages, growth of GDP, and fixed asset investment are statistically significant at the level of 1%.
    - Column-specific results highlighted:
      - After 2004, living costs (city average wages, log) strongly positively related to minimum wages with an elasticity of 2.96 (Column 1).
      - Fixed-asset investment positively related to minimum wages; lagged GDP per capita growth showed a negative effect (Column 2).
      - Size (shares) of secondary and tertiary industry positive but statistically insignificant.
      - Growth rate of the labor force and unemployment rate not statistically significant (Column 3).
    - Within-city variation of minimum wages is well explained; cross-city variation is not.
    - Conclusion: By controlling for main city-level variables (e.g., city average wages and CPI), most other variables do not add explanatory power; unexplained changes in minimum wages can be viewed as exogenous variations for individual firms.
  - Use in employment regressions:
    - In firm-level regressions, county minimum wages are used to increase variation.
    - Minimum wage as an explanatory variable in subsequent regressions includes lagged city variables (coefficients suppressed in reporting tables).

*Source: _wp14184 - 3.3    Macroeconomic Factors*

### 5.2    Neighbor County Pairs

### 5.2    Neighbor County Pairs

### Strategy and identification
- The strategy of using neighbor county-pairs is analogous to regression discontinuity design.  
- The idea is to use neighboring areas to control for unobservable factors that change over time.  
- County area varies in China. For counties where more than a few manufacturing firms are located, a typical county has an area from 20 to 2000 square kilometers.  
- Commute or migration across county borders incurs low cost for workers.  
- Contiguous regions are more likely to have similar employment trends (as argued in Dube et al. (2010)).  
- The treatment threshold is selected as thirty percent.  
- A large part of pairs are counties crossing provincial borders; significant difference in minimum wage hikes still happened in provinces such as Liaoning, Hunan, Jiangxi, and Guangdong.  
- One county may exist in different county pairs. Repeated counties are assigned with lower weights in estimations and standard errors are adjusted correspondingly.  
- For reference, the area of Hong Kong is 1,104 square kilometers.

### Rationale and advantages
- By examining within county-pair variation, the approach controls for all trends experienced by both counties in a pair.  
- When only one neighbor-pair is in the sample, the approach reduces to the typical difference-in-difference analysis for policy evaluation.  
- Controlling for neighbor county-pair fixed effects addresses positive correlation bias between unobserved regional determinants of firm employment and local minimum wage setting. Evidence: coefficients of minimum wages are more negative when pair fixed effects are included (see Main Results).

---

### Empirical model (summary)
- Dependent variable: lnL_it, the logarithm of a firm’s employment in a year.  
- Core dynamic specification includes one-period lagged dependent variable; estimation is by first-difference with GMM-style instruments (Difference GMM), instruments limited to one lag.  
- Baseline estimation equation (pooled with county-pair fixed effects):
  - lnL_it = α + β_1 lnL_i,t-1 + X_it β + X_ct β + γ lnMW_ct + μ_i + μ_p + τ_t + ε_it
  - μ_p denotes the fixed effect of county pair p (μ_p = 1 if firm i is located in one of the counties in county pair p).  
- γ measures the elasticity of minimum wages on firm employment (key parameter).  
- Firms that changed locations over the sample period were dropped.

### Variables related to firm employment (selected)
- Firm employment: L_i — reported as the average of a firm’s end-of-month number of employees in a year; end-of-year employees used to diagnose/replace suspected erroneous average employees.  
- Firm wage: W_i — firm’s total wage bill is sum of reported wages, monetary allowances, and unemployment insurance; firm wage equals total wages divided by employment. Because jointly determined with employment, firm wages are generally not used as explanatory variables.  
- Labor cost: measured jointly by county minimum wages (MW), city average wages, and industry average wages imputed from the firm sample. The log of minimum wages is the key regressor; city average wages controlled separately.  
- Other factor prices: price index of fixed asset investment at province level, price index of intermediate inputs at industry level.  
- Aggregate demand: industry output (total output at the 4-digit industry level) and city GDP per capita.  
- Price elasticity: measured by the Herfindahl index (HHI) for each industry at the 4-digit level; σ is negatively correlated with HHI (expect negative coefficient for HHI).  
- Labor income share: share of industry labor income in value of industry gross output (expect positive coefficient).  
- Productivity: profit margin as proxy; ownership categories (state, collective, private, foreign; foreign further divided into HMT and other countries); export-to-sales ratio included.  
- Lagged firm employment L_i,t-1 (instrumented by L_i,t-2).  
- Firm size: lagged annual real sales S_i,t-1.

---

### 6    Main Results

### Overview of estimation strategy for results
- Analysis period: 2000-2007.  
- Estimation proceeds in three steps: (1) average effects of minimum wage on firm employment and firm wages (2000-2007); (2) heterogeneous effects by firm characteristics (particularly firm wages); (3) additional robustness and diagnostics.  
- Neighbor-pair approach is used to control for local unobservables.

### Average elasticity and enforcement tightening (key findings)
- Interaction of minimum wage with year dummies used to study change over time.  
- Results summary (selected elasticities and tests):
  - Minimum wage elasticity for 2000-2004: -0.027 (not statistically significant).  
  - Minimum wage elasticity for 2005-2007: -0.067 (statistically significant at the 1% level).  
  - Subsample estimate for 2005-2007 (alternative specification): -0.103. Interpretation: an elasticity of -0.103 means that for a minimum wage hike of ten percent, firms tend to reduce hiring by 1.03 percent.  
  - Subsample estimate for 2000-2004 (alternative specification): 0.022 (insignificant).  
  - Using the whole sample without pair dummy variables: coefficient of minimum wages = -0.031 (significant at the 5% level).  
  - Using the model with pair dummy variables: coefficient of minimum wages = -0.103 (significant at the 1% level).  
- The change in estimated effects over time:
  - Minimum wages during 2000-2002 had insignificantly negative effects; effect turned more negative in 2003 (not coinciding exactly with the 2004 enforcement reform).  
  - After 2005, the effect became highly significant and more negative, suggesting a lagged effect of enforcement tightening (reform enacted on March 1, 2004).  
  - Division of sample into 2000-2004 and 2005-2007 is motivated by the clear difference: elasticity for 2000-2004 = -0.027 (not significant); elasticity for 2005-2007 = -0.067 (significant at 1%); difference between these two elasticities is statistically significant.

### Robustness to regional controls
- Introduction of other regional variables does not dilute the minimum wage effect for 2005-2007. Reported coefficients for the minimum wage vary from -0.059 to -0.103 across specifications for 2005-2007, all statistically significant.  
- For 2000-2004, effects remain statistically insignificant across different sets of macroeconomic variables.

### Firm-level covariate results (selected)
- Employment adjustment is slow at the firm level; lagged employment has strong explanatory power.  
- Firm size (annual sales) shows a negative relationship with employment; coefficient smaller in magnitude than minimum wage effect.  
- Ownership effects:
  - State firms are similar to foreign firms in employment levels after the tightening, but hire significantly more before 2005.  
  - Private firms generally hire fewer workers compared with foreign firms over 2000-2007.  
  - Firms with HMT investors hire more only after 2005.  
- Profitability: firms with higher profit margins tend to hire more workers.  
- Exports: firms with more exports may hire fewer workers (exports not a good proxy for labor demand in this sample).  
- HHI: negative coefficient (in line with theory).  
- Industry labor share: positive coefficient (in line with theory).  
- City average wage: negative in one specification but statistically insignificant in the presence of lagged city wages.  
- City GDP per capita: negative coefficient in one specification (not consistent with model prediction; may capture local cost factors).

### Interpretation of neighbor-pair fixed effect results
- Including neighbor-pair fixed effects yields a more negative minimum wage elasticity (-0.103) compared with the specification without them (-0.031), consistent with the argument that unobserved regional determinants of firm employment are positively correlated with minimum wage settings; controlling for these unobservables reduces upward bias.

---

### Effect of minimum wages on firm employee wages (channel evidence)
- A similar model to equation (3) is estimated with dependent variable W_t (log per-employee wage).  
- Key estimated elasticities (period-specific):
  - Minimum wage elasticity on firm wages for 2005-2007: 0.349. Interpretation: minimum hikes explain one third of firm wage increase in this period.  
  - City average wage elasticity: 0.491.  
  - Industry wage elasticity: 0.132 (2005-2007).  
  - Industry wage elasticity for 2000-2004: 0.013 (falls from 0.132 in 2005-2007).  
- Firm characteristics on wages:
  - Large firms (high sales) and private firms tend to pay lower wages.  
  - Firms with higher profitability offer higher wage rates.  
- Interpretation and implication:
  - Minimum wage hikes substantially raise average firm wages, supporting the channel through which minimum wages can affect firm employment by raising labor cost.  
  - The minimum wage has similar effect on firm wages before and after enforcement tightening, but its impact on firm employment differs across these episodes. Possible explanations include incomplete observation of fringe benefits and changes over time in firms’ ability to pass cost increases onto product prices as product market competitiveness changes.

*Italic: Source content: _wp14184 - 5.2    Neighbor County Pairs (IMF PDF chapter/section).*

### 6.2    Heterogeneous Effects

### 6.2    Heterogeneous Effects

### Grouping Based on Firm Wages
- Firms are divided into ten decile groups based on each firm’s average wage relative to the city.
- A single regression estimates the effect of minimum wages separately for these groups, controlling for the same set of other variables as in prior specifications.
- Key empirical findings:
  - Period 2005-2007 (column 1):
    - Top wage-decile firms: employment elasticity = -0.048 (statistically insignificant from zero).
    - Bottom wage-decile firms: employment elasticity = -0.153 (highly significant).
    - Interpretation: low-wage firms face binding labor demand; higher wage cost reduces labor demand and firm employment. High-wage firms may increase labor demand when minimum wages rise due to excess demand relative to labor supply under previous wages.
  - Period 2000-2004 (column 2):
    - Top decile firms: employment elasticity = 0.096 (significant).
    - Bottom decile firms: employment elasticity = -0.034 (statistically insignificant).
    - Interpretation: monotone relationship between minimum wage elasticities and firm wage persists; average effect of minimum wages on firm employment is small before enforcement tightening because firms respond less negatively or more positively.

### Full Set of Heterogeneous Characteristics
- Table X adds interaction terms of minimum wages with firm variables.
- Main results for 2005-2007:
  - Single interaction (minimum wages × firm wages) yields a positive heterogeneous effect, consistent with Table IX.
  - Adding a full set of interaction terms increases the heterogeneous effect based on firm wages from 0.117 to 0.124.
- Heterogeneity by other firm characteristics:
  - State firms tend to reduce employment much more than other firms in response to minimum wage hikes.
  - Firms with lower profit margins cut employment more than firms with higher profitability in response to minimum wage hikes.
  - Firms operating in concentrated industries do not show a tendency to reduce hirings following minimum wage increases.
- For 2000-2004:
  - The heterogeneous effect due to firm wage becomes less significant, but remains positive.
  - Heterogeneous effects due to other variables are quite similar across the two episodes.

### Robustness Checks

- 7.1 Dummy Variable of Treatment
  - Replacing continuous measurement with a dummy for large minimum wage hikes yields similar results.
  - Coefficient for the treatment dummy = -0.006.
  - Median of minimum wage hikes for treatment counties is 8 percent; columns with continuous and dummy measures are consistent, with remaining differences reflecting nonlinearity in the relationship.

- 7.2 Sample Attrition
  - Sample attrition rate ≈ 10 percent every year; re-entry rate also ≈ 10 percent.
  - Heckman selection (Heckit) used to address attrition, requiring a strong linearity assumption for the unobservable fixed effect.
  - First-stage pooled probit predictors of staying in the sample include sales, employment, ownership, profit margins, firm age, and squared firm age; county minimum wages and city fixed effects are controls.
  - Probit results (Table XII) summary:
    - Firms with larger sales, larger employment, and higher profit margins are more likely to stay.
    - Year 2004 (economic census data): coefficient of sales becomes much larger; other coefficients become insignificant.
    - Ownership: state firms most likely to leave; foreign firms most likely to stay.
    - Younger firms are more likely to stay next period.
    - County minimum wages not statistically significant except in 2003.
  - Second-stage specification adds interaction of estimated inverse Mills ratios and year dummies:
    - Estimation equation (4) specified as lnY_it = α + β1 lnY_i,t-1 + X_it β + γ lnMW_ct + μ_i + μ_p τ_t + ρ_t ˆλ_it τ_t ̄ + ε_it, where ˆλ_it τ_t is interaction of estimated inverse Mills ratios and year dummies and ρ_t is its coefficient.
  - Restricted sample: neighbor county pairs with any positive minimum wage hikes, period 2001–2005, excluding firms with hiatus periods.
  - Results (Table XIII): attrition correction barely affects estimates; inverse Mills ratios are statistically significant in years 2001, 2002, and 2003; other coefficients remain unchanged.

- 7.3 Placebos
  - Placebo test assigns treatment one year backward (pseudo treatment) for treatment counties.
  - Variable of interest: dummy indicating whether a county experiences a minimum wage hike relative to its neighbor in that year.
  - Results (Table XIV):
    - Real treatment (replicating Table VII): estimated coefficient = -0.006.
    - Pseudo treatment (one year backward): estimate = 0.003 (statistically significant).
  - Interpretation: minimum wage adjustments are hard to predict one year before, so firm expectations are unlikely to generate the real treatment effect in the placebo.

### Conclusion
- Historical and empirical summary:
  - China enacted minimum wage legislation in 1994; enforcement is commonly believed to have strengthened after 2004.
  - Using neighbor county pairs with differences in minimum wage hikes:
    - Average effects on firm employment for 2000-2002 are statistically insignificant from zero, turn negative in 2003 (rather than 2004), and become highly significant and more negative after 2005.
    - Division into episodes yields:
      - Elasticity of minimum wages for 2000-2004 = 0.022 (not statistically significant).
      - Elasticity for 2005-2007 = -0.103 (statistically significant at the 1% level).
  - Additional outcomes:
    - Minimum wage hikes increase both profit margins and employee wages in treatment counties.
    - Evidence suggests firms might not be worse off when facing an increase in local minimum wages.
- Policy-relevant interpretations:
  - Legislative tightening increased the effect of minimum wage policy, but lack of deterioration in firm profitability suggests government may have accommodated regulation adjustments to protect local firms.
  - Whether this regulatory style is welfare-enhancing or distortionary remains unclear.
  - For firms: government regulations have not clearly hindered development.
  - For employees: incumbent workers benefit from minimum wage increases; workers who quit due to policy change may be adversely affected in the short run. Long-run effects for those unemployed cannot be evaluated due to unobserved job reallocation.
  - Relevance: results shed light on later labor market regulations in China, such as the labor-contract law of 2008; labor market regulations can be binding but negative effects on firms can be mitigated by enforcement and government accommodation.

*Source: 6.2 Heterogeneous Effects (chapter/section text).*

### References

### _wp14184 - References

### Core literature cited
- Acemoglu, Daron. 2001. Good Jobs Versus Bad Jobs. Journal of Labor Economics 19(1):1–21.
- Allegretto, Sylvia A., Arindrajit Dube, and Michael Reich. 2011. Do Minimum Wages Really Reduce Teen Employment? Accounting For Heterogeneity and Selectivity in State Panel Data. Industrial Relations: A Journal of Economy and Society 50(2):205–240.
- Allegretto, Sylvia A., Arindrajit Dube, Michael Reich, and Ben Zipperer. 2013. Credible Research Designs for Minimum Wage Studies. Working Paper Series, Institute for Research on Labor and Employment.
- Arellano, Manuel, and Stephen Bond. 1991. Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations. Review of economic studies 58(2):277–297.
- Basu, Arnab K., Nancy H. Chau, and Ravi Kanbur. 2010. Turning a Blind Eye: Costly Enforcement, Credible Commitment And Minimum Wage Laws. The Economic Journal 120(543):244–269.
- Bell, Linda A. 1997. The Impact of Minimum Wages in Mexico and Colombia. Journal of Labor Economics 15(3).
- Bhaskar, Venkataraman, Alan Manning, and Ted To. 2002. Oligopsony and Monopsonistic Competition in Labor Markets. Journal of Economic Perspectives 16(2):155–174.
- Brown, Charles. 1999. Minimum Wages, Employment, and the Distribution of Income. In Handbook of Labor Economics, ed. Orley C. Ashenfelter and David Card, vol. 3, part B, 2101–2163. Elsevier.
- Brown, Charles, Curtis Gilroy, and Andrew Kohen. 1982. The Effect of the Minimum Wage on Employment and Unemployment. Journal of Economic Literature 20(2):487–528.
- Burdett, Kenneth, and Dale T Mortensen. 1998. Wage Differentials, Employer Size, and Unemployment. International Economic Review 257–273.
- Card, David, and Alan B. Krueger. 1994. Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania. American Economic Review 84(4):772–793.
- Card, David, and Alan B. Krueger. 1995. Myth and Measurement: The New Economics of the Minimum Wage. Princeton University Press.
- Casale, Giuseppe, and Changyou Zhu. 2013. Labour Administration Reforms in China. International Labour Organization.
- Draca, Mirko, Stephen Machin, and John Van Reenen. 2011. Minimum Wages and Firm Profitability. American Economic Journal: Applied Economics 3(1):129–51.
- Dube, Arindrajit, T. William Lester, and Michael Reich. 2010. Minimum Wage Effects Across State Borders: Estimates Using Contiguous Counties. Review of Economics and Statistics 92(4):945–964.
- Fang, Tony, and Carl Lin. 2013. Minimum Wages and Employment in China. Working Paper, IZA Discussion Paper.
- Flinn, Christopher J. 2006. Minimum Wage Effects on Labor Market Outcomes Under Search, Matching, and Endogenous Contact Rates. Econometrica 74(4):1013–1062.
- Freeman, Richard B. 2010. Labor Regulations, Unions, and Social Protection in Developing Countries. In Handbook of development economics, ed. Dani Rodrik and Mark Rosenzweig, vol. 5, 4657–4702. Elsevier.
- Ge, Suqin, and Dennis Tao Yang. 2014. Changes in China’s Wage Structure. Journal of the European Economic Association 12(2):300–336.
- Harrison, Ann, and Edward Leamer. 1997. Labor Markets in Developing Countries: An Agenda for Research. Journal of Labor Economics 15(S3):S1–S9.
- Hirsch, Barry T., Bruce E. Kaufman, and Tetyana Zelenska. 2011. Minimum Wage Channels of Adjustment. Working Paper, Discussion Paper series, Forschungsinstitut zur Zukunft der Arbeit.
- Hsieh, Chang-Tai, and Peter J. Klenow. 2009. Misallocation and Manufacturing TFP in China and India. Quarterly Journal of Economics 124(4):1403–1448.
- Lang, Kevin, and Shulamit Kahn. 1998. The Effect of Minimum-Wage Laws on the Distribution of Employment: Theory and Evidence. Journal of Public Economics 69(1):67–82.
- Lemos, Sara. 2004. The Effects of the Minimum Wage in the Formal and Informal Sectors in Brazil. Working Paper, IZA Discussion paper series.
- Lemos, Sara. 2009. Minimum Wage Effects in a Developing Country. Labour Economics 16(2):224–237.
- Li, Hongbin, Lei Li, Binzhen Wu, and Yanyan Xiong. 2012. The End of Cheap Chinese Labor. Journal of Economic Perspectives 26(4):57–74.
- Manning, Alan. 1995. How Do We Know That Real Wages Are Too High? Quarterly Journal of Economics 110(4):1111–1125.
- Meer, Jonathan, and Jeremy West. 2013. Effects of the Minimum Wage on Employment Dynamics. Working Paper, National Bureau of Economic Research.
- Neumark, David, and William L. Wascher. 2007. Does a Higher Minimum Wage Enhance the Effectiveness of the Earned Income Tax Credit? Working Paper, National Bureau of Economic Research.
- Neumark, David, and William L. Wascher. 2008. Minimum Wages. MIT Press.
- Rebitzer, James B., and Lowell J. Taylor. 1995. The Consequences of Minimum Wage Laws: Some New Theoretical Ideas. Journal of Public Economics 56(2):245–255.
- Schmitt, John. 2013. Why Does the Minimum Wage Have No Discernible Effect on Employment? Working Paper, Center for Economic and Policy Research.
- Stigler, George J. 1946. The Economics of Minimum Wage Legislation. American Economic Review 358–365.
- Su, Hainan, and Xia Wang. 2014. Evolution of Minimum Wage System in China and Its Effects. Research Department, Ministry of Human Resources and Social Security of China.
- Wang, Jing, and Morley Gunderson. 2011. Minimum Wage Impacts in China: Estimates from a Prespecified Research Design, 2000-2007. Contemporary Economic Policy 29(3):392–406.
- Wooldridge, Jeffrey M. 2010. Econometric Analysis of Cross Section and Panel Data. 2nd ed. MIT press.

### Selected empirical figures and descriptive trends (as reported)
- Figure II: Trend of mean city annual (effective) minimum wages and mean city nominal wages over 2000-2011 (values shown in RMB): 252, 270, 288, 305, 334, 379, 420, 485, 558, 575, 654, 790, 676, 787, 897, 1,000, 1,144, 1,290, 1,481, 1,779, 2,076, 2,309, 2,615, 2,771.
- Table I: Percentage of counties with minimum wage hikes greater than specified thresholds, sample size 2,374:
  - 1996: 36 (ą0 %), 25 (ą10%), 13 (ą20%)
  - 1997: 53, 34, 9
  - 1998: 16, 13, 4
  - 1999: 74, 71, 52
  - 2000: 26, 22, 12
  - 2001: 38, 24, 13
  - 2002: 64, 38, 12
  - 2003: 64, 27, 10
  - 2004: 75, 42, 19
  - 2005: 85, 59, 38
  - 2006: 99, 46, 18
  - 2007: 99, 65, 35
  - 2008: 99, 80, 29
  - 2009: 36, 8, 0
  - 2010: 98, 80, 11
  - 2011: 100, 97, 56
- Table II: Distribution of neighbor county differences in minimum wage hikes (2000-2007) where a hike is a neighbor difference > 1%:
  - ą0% - 10%: 58%
  - 10% - 15%: 14%
  - 15% - 75%: 28%

### Key summary statistics (Table III)
- Macroeconomic variables (prefectural cities, 1994-2011, n = 337):
  - Monthly minimum wage: Median 330; Mean 393; STD 205; Min 125; Max 1,266
  - Monthly wage per employee: Median 934; Mean 1,227; STD 862; Min 193; Max 6,319
  - Monthly GDP per capita: Median 701; Mean 1,189; STD 1,359; Min 71; Max 5,361
  - Growth rate of GDP per capita: Median 0.13; Mean 0.13; STD 0.10; Min -0.89; Max 1.21
  - Fixed asset investment to GDP: Median 0.38; Mean 0.43; STD 0.24; Min 0.01; Max 3.55
  - GDP share of secondary industry: Median 0.44; Mean 0.43; STD 0.14; Min 0.04; Max 1.00
  - GDP share of tertiary industry: Median 0.35; Mean 0.36; STD 0.10; Min 0.05; Max 1.00
  - Growth rate of FDI: Median 0.11; Mean 0.15; STD 0.67; Min -3.93; Max 10.14
  - Growth rate of labor: Median 0.01; Mean 0.05; STD 0.24; Min -2.50; Max 3.02
  - Registered unemployment rate: Median 0.03; Mean 0.04; STD 0.02; Min 0.00; Max 0.13
- Firm-level variables (manufacturing firms with consecutive presence ≥3 years, 1998-2007):
  - Employees: Median 120; Mean 285; STD 1,014; Min 2; Max 8,151
  - Sales (in thousand): Median 18,426; Mean 81,798; STD 732,096; Min 10; Max 87,000,000
  - Monthly firm employee wage: Median 967; Mean 1,205; STD 981; Min 837; Max 3,693
  - Export/Sales: Median 0.00; Mean 0.18; STD 0.36; Min 0.00; Max 1.27
  - Profit/Sales: Median 0.02; Mean 0.01; STD 0.15; Min -1.14; Max 0.38
- Firm shares over years (1999, 2001, 2003, 2005, 2007):
  - State firms: 31%, 20%, 12%, 7%, 5%
  - Private firms: 50%, 59%, 65%, 71%, 72%
  - HMT firms: 11%, 12%, 12%, 11%, 11%
  - Foreign firms: 8%, 9%, 10%, 12%, 12%

### Selected econometric estimates (preserving reported coefficients and significance)
- Table IV: Minimum Wage Accounting (dependent variable: Log(Minimum Wage), sample 346 cities, 1994-2011)
  - city average wage: coefficients reported in columns include 2.960***, 2.386***, 2.386***, ́0.210, 0.053, 0.056 (with standard errors shown in parentheses in source)
  - growth of GDP per capita: ́0.068***, ́0.069***, ́0.029, ́0.030
  - fixed asset investment: 0.069***, 0.069***, 0.012, 0.012
  - Observations across columns: 2768, 2600, 2600, 3385, 3173, 3173
  - Cities: 346, 325, 325, 346, 325, 325
  - Within R-Square: 0.88, 0.90, 0.90, 0.84, 0.85, 0.85
- Table V: Effect of Minimum Wages on Firm Employment (dependent variable: log(employees), neighbor county pairs, GMM-style instruments, 2000-2007)
  - MŴt (pă2005): 0.022, ́0.028
  - MŴt (pě2005): ́0.103***, ́0.067***
  - MŴt (pă2004): 0.011, ́0.040**
  - MŴt (pě2004): ́0.084***, ́0.053**
  - Year-specific MŴt coefficients (whole sample): 2000: ́0.014; 2001: ́0.002; 2002: ́0.025; 2003: 0.048**; 2004: ́0.045**; 2005: ́0.083***; 2006: ́0.058***; 2007: ́0.086***
  - Observations reported across columns include 1,482,311; 1,710,084; 1,147,333; 2,045,062; 3,192,395 (as applicable)
- Table VI: Effect of Minimum Wages on Firm Employment (different city controls; after vs before enforcement tightening)
  - MWt after tightening: ́0.059***, ́0.091***, ́0.103*** (columns 1–3)
  - MWt before tightening: ́0.018, ́0.022, 0.022 (columns 4–6)
  - L t-1: 0.457***–0.588*** across columns
  - sales t-1: ́0.057*** (after), ́0.051*** (before)
  - Observations: 1,710,084 (columns 1–3); 1,482,311 (columns 4–6)
- Table VII: With vs without county-pair dummy controls
  - MWt with pair dummy: ́0.103*** (column 1)
  - MWt without pair dummy: ́0.031** (column 2)
  - Observations: 1,710,084 (both columns)
- Table VIII: Effect of Minimum Wages on Firm Per Employee Wages (dependent variable: log(firm average wage), 2000-2007)
  - MWt after tightening (2005-2007): 0.349*** (column 1)
  - MWt before tightening (2000-2004): 0.329*** (column 2)
  - Observations: 1,710,084 (column 1); 1,482,311 (column 2)
- Table IX: Heterogeneous effect by firm lagged wage deciles (2005-2007)
  - MWˆfirm wage (0-10%): ́0.153*** (after), ́0.034 (before)
  - MWˆfirm wage (10%-20%): ́0.141***, ́0.027
  - MWˆfirm wage (20%-30%): ́0.132***, ́0.014
  - MWˆfirm wage (30%-40%): ́0.127***, ́0.005
  - MWˆfirm wage (40%-50%): ́0.122***, 0.003
  - MWˆfirm wage (50%-60%): ́0.116***, 0.011
  - MWˆfirm wage (60%-70%): ́0.107***, 0.023
  - MWˆfirm wage (70%-80%): ́0.099***, 0.037*
  - MWˆfirm wage (80%-90%): ́0.081***, 0.058***
  - MWˆfirm wage (90%-100%): ́0.048, 0.096***
  - Observations: 1,710,084 (after); 1,482,311 (before)
- Table X: Heterogeneous interactions (2005-2007)
  - MWt × firm wage t-1: 0.117***, 0.124*** (columns 1–2)
  - MWt × sales t-1: ́0.025***, ́0.029***
  - MWt × SOE t: ́0.087***, ́0.123***
  - MWt × profit margin t-1: 0.183***, 0.115***
  - Observations: 1,710,084 and 1,482,311 depending on column
- Table XI: Treatment variable choice
  - Treat t (binary): ́0.006*** (column 1)
  - MWt (continuous): ́0.103*** (column 2)
  - Observations: 1,710,084 (both)
- Table XII: Firm attrition determinants (pooled probit, firm stay t+1)
  - MWt coefficients by year: 2000: ́0.127; 2001: ́0.244*; 2002: ́0.096; 2003: 0.373***; 2004: 0.032
  - sales t: positive and significant across years (e.g., 0.180*** to 0.331***)
  - firm age t: negative and significant (e.g., ́0.109*** to ́0.161***)
  - Observations: 121,449 (2000); 132,301 (2001); 144,300 (2002); 159,496 (2003); 231,075 (2004)
- Table XIII: Attrition bias correction (Heckit vs No Heckit, 2001-2005)
  - MWt: ́0.103*** (both columns)
  - Observations: 1,708,758 (both)
- Table XIV: Placebo pseudo treatment years (2001-2005)
  - Treat t (standard): ́0.006*** (column 1)
  - Pseudo backward treat t: 0.003** (column 2)
  - Observations: 1,710,084 (both)

### Appendix — Variable definitions and data sources (selected highlights)
- County variable (ministry of labor)
  - County: administrative division at the third level.
  - MW: annual minimum wage; annual minimum wages are average monthly minimum wages weighted by their durations within a year. Real effective minimum wages used in firm-level regressions are denominated by the 4-digit industry product price index.
- City variables (CNKI yearbook database)
  - City: administrative division at the second level.
  - city average wage: city wage per employee (formal sector urban employees). Missing values/outliers addressed using city total wages and alternative employee sources.
  - city GDPPC: city GDP per capita; missing values/outliers replaced using GDP and annual population (annual population = simple average of beginning and end-of-year population).
  - fixed asset investment: ratio of fixed asset investment to GDP (winsorized 1% at two sides).
  - growth rate of FDI: growth rate of foreign direct investment (FDI in urban area used to replace missing values/outliers).
  - unemployment rate: registered unemployment rate (registered unemployed / urban labor force).
- Province variables (CNKI)
  - CPI: consumer price index.
  - Pk: price index of fixed asset investment.
- Industry variables (annual survey of industrial firms or other sources)
  - industry wage: real industry wage per employee (2-digit classification, GB/T4754-2002).
  - industry output: real industry output per employee (4-digit classification).
  - HHI: Herfindahl index of firm sales in an industry (4-digit classification).
  - labor shareIND: industry labor income share = total wages / total gross output (4-digit).
  - Pm: price index of intermediate input (collected by BBZ (2011)).
- Firm variables (annual survey of industrial firms)
  - Lemployees: annual firm employment (average of end-of-month employees).
  - S: sales (annual sales revenue from main business).
  - W: firm wage per employee = (wage bill + worker benefits + unemployment insurance) / annual employment; winsorized 0.5% top and dropped if < 1,000 Yuan/year.
  - Ownership dummies: HMT (Hong Kong, Macau, Taiwan), SOE (state firms), PRV (domestic private firms).
  - profit margin: profit/sales (winsorized 0.5% two sides).
  - export/sales: winsorized 0.5% from the top.

*Italic: Content synthesized strictly from the provided PDF content unit _wp14184 - References.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14184.pdf_