01        02        03       04       05       06
Restaurant Analysis · Data Analytics

Labor Market Disruption: Analyzing Policy & Economic Impacts on NY Restaurants (2016-2022)

A statistical deep-dive revealing that NY restaurants post-2020 didn't lose staff — they just got a lot more expensive to run, with labor costs surging 27%.

Market Disruption Payroll Analysis Employment Distribution Immigrant Owned Businesses Policy Impact


This project analyzes the New York State restaurant industry (NAICS 72251) to evaluate the impact of policy shifts and economic conditions between 2016 and 2022. Using Census Bureau data, the analysis investigates two core hypotheses: whether "Strict Policy" periods (2017–2020) and "Post-Policy" periods (2021–2022) fundamentally altered staffing density and labor costs.

Key Insight: While staffing levels per establishment showed a slight, statistically insignificant decline, labor costs (payroll per employee) experienced a statistically significant surge in the post-policy period, suggesting a fundamental shift in the cost of labor supply rather than a structural change in headcount.



Problem Statement

The restaurant industry faces constant volatility. This analysis aims to quantify structural changes in the labor market by answering:

  1. Did the number of employees per restaurant change significantly across policy periods?
  2. Did the cost of labor (annual payroll per employee) shift significantly?
  3. How did different legal forms of organization (e.g., S-Corps vs. Corporations) react to these pressures?



Data & Methodology


Data Source: Aggregated industry data for New York (GEO_ID: 0400000US36), specifically for "Restaurants and other eating places" (NAICS 72251).

Variables Analyzed:

  • ESTAB: Number of Establishments
  • EMP: Total Employees
  • PAYANN: Annual Payroll
  • Derived Metrics: avg_emp_per_estab (Staffing Density) and avg_payann_per_emp (Average Wages).

Period Definitions:

  • Pre-Policy (2016): Baseline.
  • Strict Policy (2017–2020): Period of heightened regulation/enforcement.
  • Post-Policy (2021–2022): Recovery and adjustment period.


    Exploratory Data Analysis (EDA)

    Initial time-series visualization revealed two distinct trends:

    1. Staffing Density: Remained relatively stable during the "Strict Policy" phase (~13.6 employees/establishment) but dipped in the "Post-Policy" phase (~11.6).
    2. Wage Inflation: Payroll per employee remained flat during the "Strict Policy" phase (~$23.5k) but spiked dramatically in 2021–2022 (~$30k).







    Statistical Analysis

    To validate the visual trends, I performed ANOVA tests and Linear Regression models.

    A. Staffing Density (Employees per Establishment)

    • Hypothesis: Policy changes caused restaurants to operate with fewer staff.
    • Result: The drop from 13.6 to 11.6 average employees was not statistically significant (p=0.171).
    • Regression: The model explained only 48% of the variance (Adj. R2=0.48), and the coefficients for policy periods were not significant at the 95% confidence level.






    B. Labor Costs (Payroll per Employee)

    • Hypothesis: Policy changes caused a significant increase in labor costs.
    • Result: The increase in wages was highly statistically significant (p=0.0099∗∗).
    • Regression: The model explained 92% of the variance (Adj. R2=0.92).

      • Post-Policy Baseline (Intercept): $30,056
      • Strict Policy Coefficient: -$6,478 (Significant difference)
      • Pre-Policy Coefficient: -$8,246 (Significant difference)






    C. Organizational & Size Analysis

    • S-Corporations showed the most volatility, dropping from ~220k employees (2019) to ~172k (2021) before recovering to ~209k (2022).
    • Wage Growth: S-Corps saw average payroll per employee jump from $24.7k (2019) to over $30k (2021).







    When analyzing the industry by scale, a clear divergence emerges. Smaller venues (1-19 employees) remained relatively stable throughout the policy periods, likely due to lower overhead and higher agility. in contrast, larger establishments—particularly those with 50 to 249 employees—experienced a severe contraction in 2021, dropping by nearly 50% in count before rebounding in 2022.







    Skewness Analysis:
    • Employment distribution was negatively skewed (-1.41), indicating a tail of smaller establishments.
    • Payroll distribution was moderately positive (0.66), suggesting wage outliers at the top end.

  • The "Wage Shock" is Real: The defining characteristic of the 2021-2022 period was a ~27% increase in labor costs compared to the strict policy period.
  • Leaner Operations: While not statistically confirmed due to small sample size (N=6 years), the data suggests a trend toward "leaner" restaurants (fewer employees per location) in the post-policy era, likely a reaction to rising wages.
  • Resilience of S-Corps: Despite a massive dip in headcount during 2021, S-Corporations rebounded quickly in 2022, proving to be a flexible legal structure in volatile times.





    • Code Snippet: Regression Analysis

    The following R code was used to determine the statistical significance of the wage increase:
    # Linear model for Payroll per Employee
pay_model <- lm(avg_payann_per_emp ~ policy_period, data = labor_trends_clean)
summary(pay_model)# Pairwise T-tests with Bonferroni correction
pairwise.t.test(labor_trends_clean$avg_payann_per_emp, 
 labor_trends_clean$policy_period,
 p.adjust.method = "bonferroni")



    Conclusion & Future Work

    The analysis suggests that the primary challenge for NY restaurants post-2020 is not a structural loss of establishments, but a rapid repricing of labor. Restaurants are paying significantly more per employee, which correlates with the observed (though statistically weaker) trend of reducing headcount per location.

    Future Considerations:
    • Inflation Adjustment: Adjusting PAYANN for CPI inflation to determine if real wages increased or just nominal wages.
    • Geographic Granularity: Breaking down the analysis by county (NYC vs. Upstate) to see if the wage shock is localized to the metropolitan area.


    Github


    References

    United States Census Bureau. (2016–2022). County Business Patterns (CBP). Data for New York State (FIPS 36), NAICS Code 72251 (Restaurants and other eating places).

    Executive Office of the President, Office of Management and Budget. (2017). North American Industry Classification System (NAICS).
    • Code 72251: Defined as establishments primarily engaged in providing food services to patrons who order and are served while seated (i.e., full-service) or order at a counter (i.e., limited-service).