Research

Regulatory Impact on Small Business Establishments

Executive Summary

American Action Forum (AAF) research examines the private sector implications of regulatory cost burdens. In particular, we analyze the cumulative effect of regulations on the number of businesses for a range of establishment sizes and find that regulatory costs have a highly regressive impact on private industries. Specifically, with a 10 percent increase in cumulative regulatory costs, there is a 5 to 6 percent fall in the number of businesses with fewer than 20 workers. That translates to a loss of over 400 small businesses in an industry. Meanwhile, those same regulations are associated with a 2 to 3 percent increase in businesses with 500 or more workers, indicating that those larger businesses are more capable of absorbing regulatory cost burdens. Small businesses will have a more difficult time complying with the cumulative effect of regulations, which could result in lost jobs.

Introduction

The interaction between regulation and private industries is highly complex. Since 2008, the federal government has imposed $733.9 billion in regulatory costs. AAF research indicates that the cumulative cost of all regulatory compliance devastates small businesses. Specifically, for every 10 percent increase in regulatory costs in an industry, the number of small and medium-size businesses in that industry falls 3 to 6 percent. The number of large businesses, meanwhile, grows 2 to 3 percent. In sum, we find that regulations cumulatively have a highly regressive effect, substantially reducing the smallest businesses and growing the largest.

Methodology

In a series of previous papers, AAF closely examined the cumulative impact of new regulations on private industries. In our last of these papers, for instance, we found statistically significant evidence that every $1 billion in new regulatory costs is associated with a 3.6 percent decline in industry-level employment.

In this paper, we study the same industries and regulations, but aim to dissect the impact of cumulative regulatory cost burdens by business size. Specifically, we estimate the relationship between cumulative regulatory costs in an industry and the number of business establishments with 1 to 4 workers, 5 to 9 workers, 10 to 19 workers, 20 to 49 workers, 50 to 99 workers, 100 to 249 workers, 250 to 499 workers, 500 to 999 workers, and those with 1,000 or more workers.

Data

To examine the effect of new regulatory costs, for each business size category we estimate the change in the number of establishments in an industry associated with an increase in the affected industry’s regulatory cost burden. AAF employs industry-level data for each business size category from the Census Bureau’s 2012 County Business Patterns and uses average number of establishments in the industries[1] in each year from 2003 to 2012.[2] The regulatory cost estimate for each industry in a year is the sum of the projected annual costs of all new regulations an industry faced from the beginning of the time to the given year. We also adjust regulatory cost projections for inflation to 2012 dollars.

Empirical Model

Using these data, AAF performs a fixed effects cubic regression to estimate the effect of an increase in regulatory costs in an industry on the number of business establishments by business size. Both the establishment and the cost terms are transformed into logarithmic variables. The cubic model with logarithmic variables allows us to address a nonlinear relationship between industry establishments and cumulative regulatory costs. In addition, we pool the business establishment data for all sizes under one business variable and use binary variables that represent each business size category. We then interact those categorical binary variables with the cost terms to estimate the association between cumulative regulatory costs and number of establishments in each business size category. For more information on our model and its exact specifications, see the appendix.

Findings

We find that regulatory costs are cumulatively associated with statistically significant changes in the number of industry establishments for each business size.

Table 1: Results

Business Size

Average Marginal Effect

1 to 4

-5.0%***

5 to 9

-5.5%**

10 to 19

-5.8%*

20 to 49

-4.0%***

50 to 99

-3.6%***

100 to 249

-2.3%**

250 to 499

-0.7%***

500 to 999

1.7%***

1000 or more

3.4%**

*Jointly Significant at the 10% Level

**Jointly Significant at the 5% Level

***Jointly Significant at the 1% Level

Average marginal effect of a 10 percent increase in cumulative regulatory cost burden on number of business establishments

 

The results in Table 1 indicate that regulations cumulatively have a highly regressive effect on businesses. While regulations cumulatively reduce the number of small and medium-size businesses, they are associated with an increase in the number of large businesses. Moreover, the results reveal that regulations harm the smallest businesses the most. In an average industry, a 10 percent increase in the cumulative cost of regulations is associated with a 5.0 percent decrease in the number of businesses with 1 to 4 employees, a 5.5 percent decrease in the number with 5 to 9 employees, and a 5.8 percent decrease in the number with 10 to 19 employees. To put these figures in perspective, an average industry in 2012 had 4,848, 1,617, and 1,311 businesses with 1 to 4 workers, 5 to 9 workers, and 10 to 19 workers respectively. If in the following years, an average industry faced a 10 percent increase in cumulative regulatory costs, it would lose 240.5 businesses with 1 to 4 workers, 88.9 with 5 to 9, and 75.5 with 10 to 19. This means the industry would lose over 400 businesses that have fewer than 20 workers.

Table 2: Implications

Business Size

Average Change in Number of Businesses

1 to 4

-240.5

5 to 9

-88.9

10 to 19

-75.5

20 to 49

-42.0

50 to 99

-7.6

100 to 249

-3.1

250 to 499

-0.4

500 to 999

0.6

1000 or more

1.5

 

Our results reveal that the number of medium-size businesses is negatively related to cumulative regulatory costs as well, although to a lesser degree. A 10 percent increase in the cumulative cost of regulations is associated with a 4.0 percent decline in the number of businesses with 20 to 49 workers, a 3.6 percent decline in the number with 50 to 99, a 2.3 percent decline in the number with 100 to 249, and a 0.7 percent decline in the number with 250 to 499. Notice that as the business size gets larger, the negative impact of regulations becomes weaker, perhaps because larger businesses are more capable of absorbing regulatory costs than smaller businesses.

To put these results in perspective, in 2012 an average industry in our sample had 1,045, 213, 132, and 53 businesses with 20 to 49 workers, 50 to 99 workers, 100 to 249 workers, and 250 to 499 workers respectively. Our results indicate that if in the following years an average industry faced a 10 percent increase in cumulative regulatory costs it would lose 42 businesses with 20 to 49 workers, 7.6 businesses with 50 to 99 workers, 3.1 businesses with 100 to 249 workers, and less than 1 business with 250 to 499 workers.

Finally, our results indicate that the number of establishments in the largest business categories actually grows when regulatory costs increase. Specifically, a 10 percent increase in cumulative regulatory costs is associated with a 1.7 percent increase in the number of businesses with 500 to 999 workers and a 3.4 percent increase in the number of businesses with 1,000 or more workers. In 2012, the industries in our sample averaged 34 businesses with 500 to 999 workers and 45 with 1,000 or more workers. Thus, the results indicate that if an average industry faced a 10 percent increase in cumulative regulatory costs, it would gain less than one business with 500 to 999 workers and 1.5 businesses with 1,000 or more workers.

Why would large businesses grow, despite facing regulatory costs? It could be that the largest businesses are most able to absorb regulatory costs, giving them a competitive advantage over smaller companies.[3] Previous research has found that this regressive trend is particularly apparent in the financial services industry. According to a 2012 Government Accountability Office (GAO) report, “[r]esearch suggests that one area in which large banks are able to take advantage of economies of scale is regulatory compliance, which contributes to their advantage in terms of operational efficiency.”

Discussion

Generally, regulatory costs are fixed, meaning that if all businesses are forced to deal with hundreds of hours of new paperwork, the costs of hiring an additional compliance officer will fall disproportionately on small institutions. Today, there are more than 236,000 regulatory compliance officers and they command average salaries of about $66,000 annually. AAF wrote on the steady growth of regulatory compliance staff, noting the “canary in the coal mine effect.” That is, we know that regulation likely increased during the recession because regulatory compliance staff grew by 18 percent between 2009 and 2012, the height of poor U.S. economic growth.

A 2013 Minneapolis Fed study emphasized the paperwork burdens and what being forced to hire compliance staff means for small banks. The study found that hiring two additional compliance officers reduced profitability by 45 basis points (roughly half-a-percent) and that one-third of the small banks studied would become unprofitable.[4] Moreover, Federal Reserve Board Governor Daniel Tarullo has warned policymakers about regressive regulation: “Any regulatory requirement is likely to be disproportionately costly for community banks since the fixed costs associated with compliance must be spread over a smaller base of assets.”

Beyond the empirical evidence that regulations disproportionately affect smaller institutions, regulators actually admit that rulemaking hurts small businesses. In a set of recently proposed efficiency standards of furnaces, the Department of Energy (DOE) noted that conversion costs for the measure would equal 18 percent of small business revenue, compared to just three percent of large business revenue. Furthermore, in EPA’s greenhouse gas regulation reporting rule, the agency noted that costs per entity for the smallest firms would be 1.32 percent, compared to 0.02 percent for the largest companies. In other words, for the reporting rule, regulatory costs are 65 times more burdensome for small businesses. Finally, in one regulation, DOE stated that the rule would likely cause small businesses to leave the air conditioning market or merge with larger competitors. DOE wrote, “It is possible the small manufacturers will choose to leave the industry or choose to be purchased by or merged with larger market players.” Regulations unquestionably have regressive impacts; just ask the regulators.  

Conclusion

Regulation does not just affect small businesses, it hurts the smallest more significantly than medium-to-large-sized establishments. The data are clear: as regulatory burdens increase the smallest businesses (1-19 employees) shed hundreds of establishments while the largest businesses (1,000 or more employees) actually grow by 3.4 percent. Despite regulatory reform designed to protect small businesses, sheer economies of scale and unchecked regulators have made life for small employers incredibly burdensome.

Appendix

As previously mentioned, AAF performs a fixed effects cubic regression to estimate the effect of an increase in regulatory costs on number of business establishments by business size. Both the establishment and the cost terms are transformed into logarithmic variables. The cubic model with logarithmic variables allows us to address a nonlinear relationship between industry establishments and cumulative regulatory costs.

We pool the business establishment data for all establishment sizes under one business variable and insert categorical binary variables that indicate which business size is being examined. For instance, when estimating the impact of regulatory costs on establishments with 1 to 4 workers, the binary variable representing that business category equals 1 and all other business binaries equal 0. Those binary variables also interact with each regulatory cost term in order to measure the cumulative impact of regulations on the number of businesses in each size category. As a result, the three cost variables (log(Cost), log(Cost)2, and log(Cost)3) and the three terms that interact each cost variable with the business size’s categorical binary variable (Business Size x log(Cost), Business Size x log(Cost)2, and Business Size x log(Cost)3) capture the total effect of regulatory costs on business establishments for any business size. For instance, the variables log(Cost), log(Cost)2, log(Cost)3, (1 to 4) x log(Cost), (1 to 4) x log(Cost)2, and (1 to 4) x log(Cost)3 capture the impact of cumulative regulatory costs on businesses with 1 to 4 workers.[5]

We use industry fixed effects to control for characteristics that vary across industries, but not over time. Also, to account for macroeconomic forces that change over time, such as the loss of businesses and jobs during the Great Recession, we control for year. To account for changes in prices during the time period, we control for industry chained Consumer Price Index. Also, any fixed effects model can face the problem of autocorrelation, in which a variable is correlated with itself over time and biases the results. Our model addresses this issue by using heteroskedasticity-and autocorrelation-consistent standard errors. The exact model is displayed below.

The subscripts i and t denote industry and year of the observations, respectively. There are three variables representing the regulatory cost burden, including log(Cost), log(Cost)2, and log(Cost)3. Those are then interacted with each binary business size variable: (1 to 4), (5 to 9), (10 to 19), (20 to 49), (50 to 99), (100 to 249), (250 to 499), and (500 to 999). CPI represents the annual average chained Consumer Price Index, as reported by the Bureau of Labor Statistics. Finally, each Yr variable is a binary variable representing the year of an observation.[6]



[1] The industries we examine can be found in “The Cumulative Impact of Regulatory Cost Burdens on Employment,” American Action Forum, May 2014, http://americanactionforum.org/research/the-cumulative-impact-of-regulatory-cost-burdens-on-employment

[2] U.S. Record Layout, County Business Patterns: 2012, Census Bureau, http://www.census.gov/econ/cbp/download/

[3] In some instances, medium businesses could be adding employees and growing into the larger business categories, resulting in a reduction in the number of medium businesses and an increase in the number of large businesses. This could be true for the 250 to 499 category, as some businesses with close to 499 workers may add more employees, resulting in fewer businesses in that category and more businesses in the 500 to 999 worker category. However, it is highly unlikely that businesses in the smaller categories are adding workers in this manner, simply because in each small business category, the relationship between regulations and establishments is negative.

[4] Marshall Lux and Robert Greene, “The State and Fate of Community Banking,” available at http://www.hks.harvard.edu/centers/mrcbg/publications/awp/awp37.

[5] To avoid the dummy variable trap, the binary variable that represents businesses with 1,000 or more workers was excluded from the model. As a result, the total effect of regulatory costs on those businesses was captured only by the three log(Cost) terms.

[6] We purposefully omit Yr1 from the regression model to avoid the dummy variable trap.

Disclaimer