Entrepreneurship and State Interventions

State intervention in labor markets, particularly through minimum wage legislation and broader institutional policies affecting economic freedom, may alter establishment dynamics differently across regions and industries. When statutory wage floors rise, establishments may reduce hiring, delay expansion, substitute toward automation, or exit the market, especially in low-wage-reliant industries and lower-income counties where prevailing wages lie closer to the statutory minimum. Because labor market conditions differ across space, uniform state policies generate nonuniform effects: Lower-income counties with lower prevailing wages are more likely to experience binding wage floors than metropolitan areas where market wages already exceed statutory minimums.

This article argues that Austrian economists should make greater use of econometric methods when the unit of observation closely approximates the actual decision-maker. County-level establishment counts provide a useful proxy for entrepreneurial responses because they reflect the cumulative outcomes of decisions regarding market entry, survival, expansion, contraction, and market exit. Econometric analysis at this level allows researchers to examine how state-level interventions shape entrepreneurial activity across heterogeneous local environments. Because policymakers increasingly rely on econometric evidence when evaluating legislation, communicating in the language of modern empirical analysis may also help advance more liberty-enhancing policy. Building on these insights, this article constructs a national county-level panel dataset to estimate how changes in state minimum wage rates and economic freedom influence establishment dynamics across industries and establishment-size categories.

Theoretical and Empirical Perspectives on State Intervention and Establishment Dynamics

Research in Austrian economics and related empirical literature suggests that state intervention can substantially influence entrepreneurial activity and establishment formation, survival, and growth (Autor et al. 2007). Austrian economists emphasize that entrepreneurship emerges through decentralized decision-making guided by market prices, profit-and-loss signals, and voluntary exchange, while institutional constraints such as minimum wages, taxes, labor mandates, and regulations may distort entrepreneurial calculation and resource allocation (Mises 2007, 1, 225–26; Rothbard 2009, 509–17; Bylund 2022).

Austrian economists view minimum wage laws as imposing a labor-market price floor above market-clearing wage rates (Mises 2007, 1, 225–26; Rothbard 2009, 509–17; Hendrickson 2019). When labor costs rise, establishments may reduce hiring, substitute capital for labor, delay expansion, or exit the market. Smaller establishments and startups are less able to absorb these increased labor costs, making it harder for entrepreneurs to scale into large enterprises.

Historically, labor-market interventions have also intersected with broader political and regional economic concerns. Thomas Leonard (2005) documents that some Progressive Era supporters of minimum wage legislation viewed wage floors as a mechanism for excluding workers they considered “unfit” for employment. Murray Rothbard (2006a, 186–87; 2006b) and Armand Thieblot Jr. (1975) document how federal minimum wage and prevailing wage legislation drew support from northern legislators seeking to blunt competition from lower-wage southern producers. Thomas Sowell (1984, 87–110) and Walter Williams (1982) further argue that prevailing wage laws disproportionately harmed lower-wage and minority workers.

These effects should be strongest in low-skilled-labor industries, where thinner margins heighten sensitivity to labor-cost increases. As minimum wages rise, managers may curtail expansion, increase automation, or consolidate more aggressively than higher-skilled-labor establishments. Austrian economists further argue that such policies disproportionately harm establishments employing inexperienced workers, immigrants, minorities, and younger employees while favoring larger incumbents (Bellante 2007; Hendrickson 2019). David Audretsch and Max Keilbach (2004) identify “entrepreneurship capital” as a key regional asset, and Simon Parker (2018, 654–755) shows that economic and political institutions shape establishment formation.

Broader differences in economic freedom may reinforce these pressures. Austrian economists argue that higher taxes, regulatory burdens, occupational licensing requirements, labor mandates, and restrictions on voluntary exchange increase the fixed and variable costs of operating a business (Rothbard 2009, 875–1073). Establishments dependent on low-skilled labor are especially vulnerable, facing higher turnover, greater regulatory exposure, and stronger price competition. As compliance costs rise, entrepreneurs may hire and invest less, leading establishments to stay small, automate, relocate, or exit. Empirical research similarly finds that establishments shift economic activity across jurisdictions in response to regulatory variation (Weng and Peng 2018). Spatial equilibrium models predict that uniform regulatory floors bind more tightly in lower-income areas (Roback 1982).

Methodology

This article estimates a dynamic two-way fixed-effects panel model to assess the impact of two primary policy variables—the real effective minimum wage and the Economic Freedom of North America subnational index—on establishment counts across US counties. The analysis produces two main sets of results: (1) establishments across all industries and (2) establishments in low-wage-intensive industries. Within each set, separate equations are estimated for four establishment size classes (1–9, 10–19, 20–49, and 50–499 employees), with the baseline specification as follows:

$$\begin{aligned} y_{kit} &= \alpha + \beta_{k}\ln\left( {mw}_{it - 1} \right) + \theta_{k}\ln\left( {efsi}_{it - 1} \right)\\ & \quad + \mathbf{\gamma}_{k}\mathbf{\mathstrut'}\mathbf{W}_{it - 1} + \mathbf{\delta}_{k} \mathbf{\mathstrut'} \mathbf{X}_{it} + \lambda_{i} + \tau_{t} + \phi_{t} + \varepsilon_{kit}, \end{aligned}$$

where k indexes establishment size categories, i indexes counties, and t indexes years. The dependent variable y_kit is the number of establishments in size class k in county i at time t, in natural logarithms.

The two policy variables of interest are the natural logarithms of the effective state-level minimum wage rate (mw) and the Economic Freedom of North America subnational index (efsi). These variables are included in lagged form for several reasons. Establishments adjust to policy and economic conditions with delay due to planning horizons, contractual rigidities, and fixed costs. Policy variables such as minimum wages may also respond to contemporaneous economic conditions. Lagging these variables ensures that they are predetermined with respect to current shocks. The coefficients β_k and θ_k measure the elasticities of establishment counts with respect to the minimum wage and economic freedom, respectively.

Several contemporaneous economic controls (W), including total establishment counts across all industries, the county unemployment rate, and real per capita personal income, may lie on the causal pathway between the policy variables and establishment outcomes. These variables are therefore entered in lagged form (in natural logarithms) as well to mitigate attenuation bias. This lag structure is applied consistently across all specifications.

Demographic and structural controls (X), such as educational attainment, poverty rates, and racial/ethnic composition, enter the model contemporaneously. Because these variables change slowly and are unlikely to respond immediately to policy shifts, they serve as appropriate controls for local conditions. The control set remains fixed across all specifications to ensure comparability.

The model includes county fixed effects (λ_i) to control for time-invariant county characteristics such as geography, long-run industrial composition, and persistent demographic factors. Year fixed effects (τ_t) absorb national macroeconomic shocks and business-cycle fluctuations. In addition, state-specific linear time trends ($\phi_{t}$) control for gradually evolving state-level factors such as long-term policy shifts, demographic changes, and industry trends. The error term ($\varepsilon_{kit}$,) captures the influence of all other factors not explicitly included in the model, after accounting for state trends, county and year fixed effects. Identification of parameters relies on within-county variation, net of national and state-specific trends. Because minimum wage policy varies at the state-year level, standard errors are clustered at the state level.

All equations share a common specification, differing only in the dependent variable (i.e., establishment counts by establishment size category and sectoral grouping). This consistency ensures that differences in estimated coefficients reflect underlying economic variation rather than differences in model design. The primary results focus on low-wage-intensive industries, where Austrian theory predicts the strongest effects from minimum wage increases and differences in economic freedom. These results are compared with parallel specifications using establishment counts across all industries as a benchmark to assess whether policy effects are concentrated in low-wage sectors or economy-wide. Comparisons across establishment size categories also reveal whether smaller establishments are more sensitive to policy changes than larger establishments.

Data

This article utilizes a comprehensive county-year panel dataset to examine the impact of changes in the effective minimum wage on establishment formation, survival, and growth. While using broader establishment counts as a benchmark, the analysis focuses on how minimum wage increases affect establishments in industries that are particularly reliant on low-skilled, low-wage labor. The dataset covers nearly all US counties from 2010 to 2024 and is summarized in table 1. Counties are identified using five-digit Federal Information Processing Standards (FIPS) codes, which are constructed by concatenating the two-digit state FIPS code with the three-digit county FIPS code. Establishment-level data are obtained from the US Census Bureau’s (2010–24) County Business Patterns, which provide annual establishment counts by North American Industry Classification System (NAICS) industry and employment size category.

The dependent variables consist of establishment counts across four establishment size categories: 1–9 (n1t9), 10–19 (n10t19), 20–49 (n20t49), and 50–499 (n50t99) employees. These counts are constructed in two ways. The first is a benchmark set (all) which includes establishments across all industries in the economy. The second set (mwF) focuses on industries that are highly dependent on low-wage labor, identified using the following NAICS classifications: full-service restaurants (722511), limited-service restaurants (722513), cafeterias and buffets (722514), snack and beverage bars (722515), clothing stores (4481), shoe stores (4482), gasoline stations (4471), grocery stores (44511), and convenience stores (44512). These sectors are characterized by labor-intensive production, relatively low capital requirements, thin operating margins, and a high concentration of workers earning wages at or near the statutory minimum. Most entrepreneurial activity occurs within the smallest size category (1–9 employees), while the larger size classes more closely reflect establishment survival and expansion. According to table 1, on average, counties contain 1,840.5 establishments with 1–9 employees across all industries and 133.1 such establishments within minimum wage–sensitive sectors. In both categories, the average number of establishments declines as establishment size increases.

The key independent variables are the real effective state-level minimum wage (mw_r) and the Economic Freedom of North America subnational index (efsi). The minimum wage variable is defined as the greater of the federal or state statutory minimum wage and is adjusted for inflation using the Consumer Price Index. As reported in table 1, the mean real minimum wage in the sample is $9.80. The economic freedom index is constructed on a scale of 0 to 10, where higher scores indicate greater economic freedom. For the sample used in this article, the index ranges from 1.5 to 8.1 with a sample mean of 6.4.

County-level control variables include real per capita personal income (pcpi_r); unemployment rate (unemp); poverty rate (pov_rate), racial and ethnic composition, represented by the share of the population that is white, black, or Hispanic (wht_rate, blk_rate, and hsp_rate); total population (pop), Hispanic population (pop_hsp), and measures of educational attainment, including the share of the population with a bachelor’s degree or higher (bs_plus) and the high school dropout rate (hs_drop). These variables are obtained from the Bureau of Economic Analysis (2010–24) and the US Census Bureau’s (2009–24, 2010–24) American Community Survey and County Business Patterns.

Analysis and Discussion

Tables 2 and 3 report the benchmark and primary regression results from the dynamic two-way fixed effects panel models. Table 2 presents the benchmark estimates using establishment counts across all industries, while table 3 reports the primary results for industries that are highly dependent on low-wage labor.

The benchmark results in table 2 indicate that minimum wage increases have little effect on establishment counts (L.ln_mw_r) across the economy as a whole, except among larger establishments with 50–499 employees, where the estimated effect is negative and statistically significant. By contrast, the estimated coefficients on the economic freedom index (L.ln_efsi) are positive across all establishment-size categories and become economically large and highly statistically significant among establishments with 20–49 and 50–499 employees. These results suggest that differences in economic freedom may matter more for establishment expansion and the survival of medium-sized establishments than minimum wage policy alone when examining all industries collectively. The findings are consistent with the Austrian view that lower taxes, fewer regulatory burdens, and greater freedom of exchange strengthen the incentives and flexibility necessary for entrepreneurial growth and business expansion.

Our primary results in table 3 reveal large, negative, and highly statistically significant minimum wage elasticities (L.ln_mw_r) across all establishment-size categories within low-wage-intensive industries. These findings suggest that minimum wage increases disproportionately affect industries that rely more heavily on low-skilled labor, consistent with Austrian theories emphasizing the role of relative prices, entrepreneurial calculation, and heterogeneous establishment responses to state intervention. The estimated coefficients on the economic freedom index (L.ln_efsi) are positive, large, and statistically significant across all four establishment-size categories, indicating that greater economic freedom is strongly associated with higher establishment counts within low-wage-intensive sectors.

Taken together, the results in tables 2 and 3 suggest that both minimum wage policy and economic freedom matter substantially more in labor-intensive industries than in the broader economy. Whereas only one of the four minimum wage coefficients in the benchmark specifications of table 2 is statistically significant, all four are negative, economically large, and statistically significant in table 3. Similarly, although the economic freedom coefficients in table 2 are relatively small and become statistically significant primarily among larger establishments, the corresponding coefficients in table 3 are substantially larger across every establishment-size category. These comparisons indicate that labor-intensive sectors are considerably more sensitive to both labor-market interventions and broader institutional conditions than establishments in the aggregate economy.

The remaining control variables generally exhibit signs and magnitudes consistent with theoretical expectations, providing additional support for the specification. Lagged establishment counts (L.ln_n_all) are positive and highly statistically significant across all equations, indicating substantial persistence in establishment dynamics and supporting the inclusion of a dynamic adjustment process. The unemployment rate exerts larger and more consistently negative effects (L.ln_unemp) within the low-wage-intensive specifications of table 3 than in the benchmark results of table 2, suggesting that labor-intensive establishments are more vulnerable to deteriorating labor-market conditions. Real per capita personal income (L.ln_pcpi_r) is generally positive and statistically significant in the benchmark models in table 2, consistent with stronger local demand supporting broader business activity, while its weaker and occasionally negative effects for larger establishments in table 3 suggest that rising incomes may also reflect higher labor costs and competitive pressures within low-wage-intensive sectors. Educational attainment also differs across the two tables, with the share of the population holding a bachelor’s degree (ln_bs_plus) generally positive or weakly related to overall establishment counts in table 2 but negative and highly significant in table 3, consistent with low-wage-intensive industries being less concentrated in highly educated labor markets. Overall, the consistency of the key coefficients, the strong persistence effects, and the differing responses across industries and establishment-size categories suggest that the model is capturing meaningful economic relationships rather than random statistical associations.

Conclusions

The results suggest that state interventions, particularly minimum wage increases and reductions in economic freedom, disproportionately affect establishments that rely heavily on low-wage labor while exerting far weaker effects across the broader economy. Because these establishments often provide entry-level employment opportunities for younger, lower-skilled, inexperienced, immigrant, and minority workers, reductions in establishment formation, survival, and expansion may make obtaining a first job more difficult, limiting these groups’ accumulation of work experience and long-run labor-market advancement. The findings also suggest that these policies disproportionately burden smaller labor-intensive establishments, increasing the likelihood of delayed expansion, automation, consolidation, or market exit.

More broadly, this article demonstrates why Austrian economists should make greater use of econometric methods when the unit of observation closely approximates the underlying decision-maker, in this case county-level establishment counts reflecting entrepreneurial decisions regarding entry, survival, expansion, contraction, and exit. By combining Austrian theory with county-level panel econometrics, researchers can communicate in the empirical language policymakers increasingly rely upon while providing evidence for how interventionist policies affect entrepreneurial activity and economic mobility.