One size fits all? The differential impact of federal regulation on early-stage entrepreneurial activity across US states

Numerous studies evaluate how a jurisdiction’s institutional and specifically regulatory environment impact firm formation and entrepreneurial activity. This study adds to this by employing a dataset measuring the differential impact that federal regulations have on industries across US states. Specifically, the paper addresses how such differences affect several facets of early-stage entrepreneurial activity, including an index measure of early-stage entrepreneurship, opportunity entrepreneurship, job creation at startup, and new firm survival rates all derived from the Kauffman Index of Entrepreneurial Activity. Overall, the results suggest that early-stage entrepreneurial activity tends to be negatively correlated with a relatively more burdensome federal regulatory environment. While the channels do not indicate an effect through new firm formation or firm survival rates, both opportunity entrepreneurship and job formation are negatively and significantly affected. Implications are discussed.

activity and development in general. Thus, it is the "rules of the game" and existing institutional structures and frameworks that will either incentivize or disincentivize individuals from pursuing such activities (See North, 1989 on the relationship between institutions and economic growth). 1 Importantly, it may not be so much that the institutional environment will impact the stock of entrepreneurs and entrepreneurial effort, however it does directly impact the type of entrepreneurial effort put forth-relatively speaking (Baumol, 1990).
Specifically, entrepreneurial efforts and opportunities are those that can be productive (and growth enhancing), unproductive (leading to net transfers, with no real accompanying growth), or destructive (which actively undermines existing capital stocks) (Baumol, 1990).
The institutional environment, then, will directly influence how entrepreneurs-both existing and would-be-devote their resources and efforts. It is through these mechanisms that the institutional environment can affect entrepreneurial efforts and with it economic growth in general.
Tied to this is the regulatory environment that exists within a given jurisdiction, how it is enforced, and how the relative burden falls on various groups or industries within an economy. Where market failures exist and result in an inefficient allocation of resources, then public regulation can be applied to correct those failures. This would promote economic efficiency thereby directing entrepreneurial endeavors toward efficient, growth-enhancing activity. On the other hand, it is possible-be it through rent-seeking, regulatory capture, or some other means-for regulation itself to generate greater inefficiencies and impose inefficiently high costs on pursuing productive and growth-enhancing economic activities. This study develops these considerations by evaluating how or even if the relative burden that federal regulations impose ultimately impact entrepreneurial efforts across all 50 US states. Given geographical, jurisdictional, and market concentration differentials, federal regulations do not tend to have a uniform impact across states. Thus, a federal regulation that uniformly applies across all states may have a disproportionate impact within those states, which may also differentially impact entrepreneurial activity across states.
To evaluate these possibilities, I use a recently developed measure of federal regulations' impact across states: the FRASE index (to be discussed in greater detail below). This measure is evaluated against various measures of early-stage entrepreneurship specifically taken from the Kauffman Index of Entrepreneurship. Included here, along with an index measure of early-stage entrepreneurship, this study also considers how rates of new entrepreneurship, opportunity entrepreneurship, early-startup job creation, and early-startup survival rates might be impacted by the relative burden that federal regulations have across US states with data from 1998 to 2017. Overall, after controlling for a number of factors, the results suggest that opportunity entrepreneurship along with early-stage job creation, and early-stage entrepreneurship as a whole are all negatively associated with disproportionately more burdensome federal regulatory environment. Finally, the rate of new entrepreneurship and early startup survival rates do not show a strong association with the federal regulatory burden in a given state. Implications are discussed.
The paper is structured as follows: a review of the relevant literature and theoretical considerations are presented in Sect. 2. A detailed discussion of the data employed and empirical specifications are provided in Sect. 3. The results and possible explanation of the results are found in Sect. 4, while Sect. 5 contains concluding remarks.

Review of the relevant literature and theory
A significant body of literature has emerged evaluating the factors that affect entrepreneurial activity as it relates to the institutional environment within a given jurisdiction. 2 Generally, it's the institutional environment that creates the particular set of "rules of the game" that individuals will face which ultimately play a primary role in fostering or hindering economic growth and development (Acemoglu & Johnson, 2005;Acemoglu et al., 2002;North, 1990;Rodrik et al., 2004;Weingast, 1995). There is also a direct link between institutional quality and entrepreneurial endeavors (whether they be productive or unproductive a la Baumol, 1990), which have been shown to generate greater economic opportunity and with it enhance growth and development (Aidis et al., 2012;Bjørnskov & Foss, 2008;Dove, 2015;Hafer, 2013;Hall & Sobel, 2008;Laeven & Woodruff, 2007;Nyström, 2008;Troilo, 2011). Much of this research suggests a strong link between increased productive entrepreneurial pursuits when property rights are protected and enforced and public predation is minimized.
From this broad-based institutional view, numerous studies have also identified links between relative regulatory burdens and how these might impact both entrepreneurial endeavors and economic growth in general. Such regulatory burdens can delay the process to starting a business and, at the margin, outright deter entry (Djankov et al., 2002;van Stel et al., 2007;Klapper et al., 2006;Chambers and Munemo 2019). Clearly, regulation exists to correct the externalities that can emerge from various market failures. However, there is also evidence of a link between decreased entrepreneurial activity and with it reduced growth as regulation increases beyond some optimum level (DeSoto, 2000;Djankov et al., 2002;Dreher & Gassebner, 2013;Klapper et al., 2006). Additionally, a shift in regulatory activity may also result in a reallocation of entrepreneurial talent from productive and growth-promoting entrepreneurship to unproductive and zero-sum entrepreneurial endeavors, if regulation (both actual and potential) shifts incentives to pursue more rent-seeking behavior, engage in regulatory capture, or rent extraction by public officials (Coyne et al., 2010;Djankov, 2009;McChesney, 1987). While regulation is meant to reduce transactions costs involved in pursuing entrepreneurial activity and promote economic growth, regulation can also become a "grabbing hand" of sorts and used to extract resources (Djankov et al., 2002), potentially even reducing the available stock of entrepreneurial talent.
Such a regulatory environment may increase the cost of pursing productive entrepreneurial activity, especially for new entrants. Thus, regulation-rather than promoting economic growth and efficiency-instead creates artificial barriers to entry and other inefficient and onerous costs. On net then, especially within jurisdictions and industries that are disproportionately affected, firm entry and formation, along with entrepreneurial efforts may be hindered and with that growth may be retarded. A strand of the literature considers just how a jurisdiction's regulatory environment, especially entry regulations, impact various aspects of entrepreneurial activity. Aidis et al. (2012) evaluate how several institutional features, including regulatory environments, affect decision-making processes about pursuing entrepreneurial endeavors across countries. The authors find that size of government has a direct effect on entry, though corruption is only weakly associated. In a similar vein to the current study Demirguc-Kunt et al. (2006) and van Stel et al. (2007) consider how a country's regulatory environment impacts rates of incorporation and nascent entrepreneurial activity respectively.
In both papers, the findings suggest that more heavily regulated economies, and especially those economies that restrict corporate entry (Demirguc-Kunt et al., 2006), and regulate capital requirements and labor markets (van Stel et al., 2007) all reduce entrepreneurial entry. Further, Dreher and Gassebner (2013) evaluate the relationship between corruption and regulation, finding that a more burdensome regulatory environment tends to hinder firm formation and entry. 3 The channel is through procedural requirements to start a business and minimum capital requirements (similar to van Stel et al., 2007), though this effect is mitigated as corruption increases. Finally, Prantl and Spitz-Oener (2009) find that entry regulations tend to deter self-employment and new business formation.
Also, tied to the current study, Klapper et al. (2006) estimate how regulatory requirements impact the formation and entry of limited liability companies (LLCs). Overall, the authors find evidence that regulatory burdens decrease new firm formation, result in relatively larger firms that do enter the market, and are associated with slower growing incumbent firms. Capelleras et al. (2008) corroborate these results and evaluate how regulation impacts the size and growth of new startups. Overall, the authors find that in relatively less regulated economies, newly registered firms tend to be smaller but grow more quickly, while greater regulatory burdens reduce the proportion of officially registered firms (but not necessarily total new firms). Chambers and Munemo (2019) also consider how entry regulations affect LLC formation across countries, finding that such formation is relatively lower as entry regulations and barriers increase. Additionally, Levie and Autio (2011) look at how various regulatory frameworks (include the regulation of firm entry, labor market regulations, and exit regulations) along with measures of "rule of law" affect entrepreneurial entry into the marketplace. Here, laxer regulatory environments tend to increase the rate of strategic entrepreneurial entry, with strong "rule of law" somewhat dampening this effect.
Greater regulatory entry hurdles also tend to hinder entrepreneurial entry and exacerbate inequality, at least across US states (Chambers & O'Reilly, 2022). Within the US, an increased regulatory burden decreases firm formation (Bailey & Thomas, 2017) and for those firms that do enter, results in lower rates of employment (Chambers and Guo forthcoming). However, this former result is somewhat contradicted by Goldschlag and Tabarrok (2018) who do not find a clear association between federal regulations and general economic dynamism. Chambers et al. (forthcoming) address these contradictory results and suggest, with evidence, that they are largely driven by differences in the dependent variables employed across studies.
The current study adds to all of this literature in several important ways. First, I evaluate the relative impact of federal regulation across US states. While not as generalizable as international and cross-country studies, evaluating US states negates the heterogeneity that can exist in cross-country analyses and thus the need to control for various additional factors. Further, regulation may be endogenous with entrepreneurship. However, this effect should be less a concern when evaluating regulation set at the federal level and entrepreneurial activity across states. From this and the literature above, I hope to test then how the relative impact of federal regulation affects various aspects of entrepreneurial (and especially early-stage entrepreneurial activity). First, based on Baumol's (1990) insights, it may not be the case that regulation will impact the stock of entrepreneurs, just the relative efforts made by those individuals (whether they pursue productive or unproductive activities). On the flipside, Djankov et al. (2002) suggest that regulatory costs negatively impact in such a manner so as to reduce the number of entrepreneurial endeavors (and thus business startups) undertaken.
As also noted above, when firms do enter they tend to be larger. This is especially so where entry barriers and labor markets tend to be more heavily regulated. Therefore, I expect that states that bear a larger relative burden of federal regulations will see relatively large firms entering the market. Specific to this study, I also anticipate that greater regulatory burdens will result in lower job creation at startup. Bertrand and Kramarz (2002) find evidence of slower employment growth in France as a result of increased entry and product regulation. Lucas and Boudreaux (2020) evaluate federal regulations' influence on job creation at the loca level within the US and find that increased regulatory burdens negatively effect net job creation, though greater economic freedom at the state level tends to moderate this effect. Along with these observations, this study also considers how the relative burden borne by states impacts growth-enhancing, opportunity entrepreneurship. Here, if regulation increases compliance costs, and makes market entry and job creation more expensive, then this may drive individuals into entrepreneurship out of necessity, rather than opportunity. As such, I also anticipate that as the regulatory burden increases within a state, this will lead to relatively lower rates of opportunity entrepreneurship within a given state. Lusardi (2009, 2010) find exactly this effect as do Aparicio et al. (2016). Regulatory burdens-especially those that effect market entry-tend to dampen opportunity entrepreneurship, and increase self-employment in relatively low-paying, low-skilled sectors especially for women and minorities (Ardagna & Lusardi, 2009). 4 Dove (2020) finds a similar result at the US state level. The next section tests these propositions along with providing a detailed description of the data and empirical specifications employed.

Data and model specification
Data for this study come from several sources. The main independent variable of interest is the FRASE Index, which is developed by QuantGov. 5 This is a normalized measure of the impact that federal regulations have on each individual state relative to the nation as a whole. This allows for consistent comparisons across states over time. As Chambers and Broughel (2022) note, the index weights how strict federal regulations are by industry, relative to a particular industry's overall share of private sector output by state and year. This is then applied to determine the ratio of a given industry's output relative to private sector output nationally. The final index is normalized by the sum of federal regulations that existed in 1997. These are then converted into normalized scores relative to the economic impact that federal regulations have on the US economy as a whole.
Thus, the measure can be thought of as a ratio, which considers the impact of a federal regulation on the private sector of each state relative to its impact on the private sector nationally. There are two particular scores computed: the constant-basis score and current-basis score. The current-basis score provides a measure for each individual year available, while the constant-basis score provides a score for the cumulative effect (over all years available) that federal regulations have on each state's private sector relative to the nation as a whole. Data are available from 1997 forward. 6 To get a sense of magnitudes and what each score represents, a score of "1" within the FRASE index would indicate that the impact that federal regulations have on a state's private sector is equivalent to its effect on the private sector for the nation as a whole. A score of "1.5" would suggest that a state's private sector is impacted 50% more relative to the nation as a whole, while 0.75 would indicate that a state's private sector is impacted 25% less than the nation as a whole.
The dependent variables employed are derived from the Kauffman Index of Entrepreneurial Activity. These measures evaluate several facets of entrepreneurship and entrepreneurial endeavors across each state over numerous years. Here, this study uses several measures of early-stage and early-startup entrepreneurial activity. The Kauffman Early-Stage Entrepreneurship (KESE) Index is a normalized composite score composed of four subcomponents (to be discussed individually in greater detail below) with a mean of zero, and which applies equal weighting to all subcomponents. This measure thus provides a general overview of early-stage entrepreneurship, which reflects early entrepreneurial activity and general trends in the first-year of business activity. 7 The analysis first considers how federal regulation impacts this KESE Index. As Fairlie (2022) notes, it is possible that KESE scores may be skewed by a single high or low score within an individual subcomponent. Therefore, along with this overall composite index, the current study also disaggregates the KESE into its four subcomponents and evaluates how these might be impacted by federal regulation.
The subindices include the following measures: Rate of New Entrepreneurs, Opportunity Share of New Entrepreneurs, Startup Early Job Creation, and Startup Early Survival Rate. Per Fairlie (2022) the rate of new entrepreneurs is a measure that includes all business owners regardless of growth potential, owner intentions, business size, business type (incorporated or unincorporated business), and whether or not they employ anyone else. Opportunity share disentangles those entrepreneurs who are not pursuing entrepreneurial opportunities due to unemployment (necessity entrepreneurship). This measure is specifically the percent of all new entrepreneurs who were not unemployed at the time of starting a new business.
Startup early job creation is a normalized (by population) measure of the jobs created by firms within their first year of operation. As it is normalized by population, it is best viewed as the jobs created within the first year of operation for an average new firm per every 1000 people, and therefore representing the job creation of a typical startup firm. Finally, startup early survival rate is a measure of new establishments still active and in operation after one year.
From this, the baseline empirical model takes the following form: where Entrepreneurshi p it represents each measure of entrepreneurship as outlined and discussed above. F R ASE it represents the constant-basis score also discussed above. X i t is a matrix that includes all of the of socioeconomic, demographic, and institutional variables meant to control for other factors influencing entrepreneurial activity across states, while μ i are state fixed effects and γ t represent times dummies. Along with these main variables of interest, there are several socioeconomic and demographic control variables included that are typical to the literature (Fritsch & Storey, 2014 provide a relevant literature review). These include the median age by state, the male and white population shares, population density, the population share over age 65, the percent of the population with a bachelors degree or more (over age 25), unionization rates by state, poverty rates, and finally a measure of state-level economic freedom. In general, an older population is negatively associated with entrepreneurship so both median age and population share over 65 are expected to be negative. The male and white population shares are expected to have an ambiguous effect, while population density tends to be positively associated with more entrepreneurial activity in general. Finally, educational attainment and greater economic freedom both tend to be positively correlated with entrepreneurship, while unionization and poverty are both negatively associated with entrepreneurship. Table 1 provides the summary statistics for all of these variables.
Unionization data are taken from the Union Membership and Coverage Database, estimated with data from the US Census Bureaus' Current Population Survey. This database includes private and public sector union membership, beginning in 1983 for each state. 8 Economic freedom scores come from the Fraser Institute's Economic Freedom of North America Index, which is compiled annually and ranks economic freedom in each state based on equally weighted subcomponents that are normalized to scores between "0" (least economically free) to "10" (most economically free). 9 All remaining variables are taken from US Census data (linearly interpolated when only decennial data are available), with educational attainment data supplemented with data from the American Community Survey over various years. Results and a detailed discussion of those results are presented in the next section.

Results and interpretation
The impact that federal regulations have on entrepreneurial activity are first presented in Table 2. The KESE Index, compiled from all four subindices discussed above, is utilized first as the dependent variable in Table 2. 10 Column 1 does not include any control variables. Column 2 includes demographic and age variables along with population density. Column 3 excludes the controls from column 2 but includes educational attainment, unionization and poverty rates, and economic freedom scores. Column 4 reports results utilizing all above mentioned control variables. Overall, there are some interesting trends that emerge. The results indicate a negative and statistically significant relationship (at the 5% level or better) with the KESE index. While the construction of both the KESE and FRASE indices make interpretation of the raw coefficient difficult, a one standard deviation increase in the Constant score decreases the KESE index anywhere between 0.17 and 0.20 standard deviations. The result would suggest that it is the gradual buildup and accumulation of federal regulations over time that has a significant effect on early-stage entrepreneurs and entrepreneurial activity.
The next several tables present the disaggregated subcomponents of the KESE index (again, the rate of new entrepreneurs, opportunity entrepreneurship, early startup job creation, and early startup survival rates). Table 3 is the first of these, which includes the rate of new entrepreneurs as the dependent variable. The layout for Table 3 follows that of Table 2 as discussed above. Here across specifications, while the coefficients are positive, none are statistically significant. The magnitudes on the coefficients remain quite small as well. Overall, these results seem to be in line with Baumol's (1990) argument that it's not so much the institutional environment that affects the stock of available entrepreneurs, but rather the relative payoff available from pursuing various types of entrepreneurial opportunities. To the extent that the rates of new entrepreneurial activity remain fairly unchanged from regulation, this might suggest that individual's shift the type of entrepreneurial endeavors they pursue (potentially from relatively more productive to unproductive opportunities as regulation shifts relative costs). However, it is difficult to disentangle this specific effect given the construction of the data.
Next, the impact that federal regulations have on opportunity entrepreneurship is explored. These results can be found in Table 4. Dove (2020) provides an initial evaluation of federal regulation's impact on opportunity entrepreneurship across states, also using the FRASE Constant Index. Expanding on this previous work, the current study adds two additional years of data to the analysis. Overall, the findings here confirm Dove's (2020) earlier estimates with all results negative and significant at the 1% level or better.
The magnitudes of the effect are also fairly large, with a one standard deviation increase in the Constant score associated with a 0.20 to 0.34 standard deviation decrease in opportunity entrepreneurship again depending on specification. Where opportunity entrepreneurship is best thought of as that entrepreneurship which emerges Robust standard errors in parentheses. Constant suppressed from output tables ***p < 0.01, **p < 0.05, *p < 0.1 as a result of pro-cyclical macroeconomic activity and general economic optimism, and is also thought to be growth enhancing entrepreneurial activity, then these results indicate that as the relative impact of federal regulations borne across states increases, this has a dampening effect ultimately on economic growth. On the flipside, the results also suggest that as a state's regulatory burden increases, it drives more individuals to pursue entrepreneurial endeavors out of necessity. These findings tend to corroborate those found within the broader literature as noted in Sect. 2. Several possible explanations based on the results may be the increased uncertainty that a changing regulatory environment might generate within a given year and also the increasing costs that such regulatory changes might impose on would-be entrepreneurs seeking growth-enhancing opportunities over time. It may also make it more difficult for individual to find gainful employment and thus drive greater self-employment out of necessity.
Next, the effect of federal regulation on early startup job creation is presented within Table 5, again with the format following that of the previous tables above.
Again, the early startup job creation variable is a measure of the total number of jobs that a new startup creates within its first year per 1,000 people. Here, there is a strong negative association between this measure of job creation and regulation. Additionally, all specifications are statistically significant at the 5% level or better. A one standard deviation increase in the constant index indicates a 0.11 to 0.17 standard deviation decrease in job creation depending on specification. These results suggest that when individuals do pursue entrepreneurial endeavors, that job creation at inception and within the first year of operation tends to be lower in those states that face a relatively larger regulatory burden compared to other states as federal regulations filter through the economy.
Finally, Table 6 lays out results when evaluating how the relative impact of federal regulations impact the survival rates of early startups. Again, this measure is simply the proportion of new firms and startups that still operate after one year.
Overall, though every specification once again nets a negative result, they are not particularly robust. Specifically, while the results for the constant index are statistically significant in 3 of 4 specifications, two of those are only marginally significant (at the 10% level) and results from the full model (column 4) are insignificant. Here, a one standard deviation increase in the Constant index decreases survival rates from 0.12 to 0.17 standard deviations.
Taken as a whole and to the extent any causal effects can be inferred, the above results point to the relative impact that federal regulations have across states being generally negative when it comes to early-stage entrepreneurship and entrepreneurial activity on net. Overall, while the rate of new entrepreneurs is not particularly affected (which potentially lends support to Baumol's (1990) insight that the institutional environment does not affect the stock of entrepreneurial talent, just the relative payoff of pursuing various types of entrepreneurship), there does appear to be a significant and deleterious effect on growth-enhancing, opportunity entrepreneurship. Further, while early-stage firm survival rates also do not tend to be particularly impacted by the overarching regulatory burden, early-stage job creation is adversely impacted. This lends support to the literature that finds a relatively more burdensome regulatory environment leads to slower firm growth for those firms that do enter the market. Along with potentially providing evidence of Baumol's (1990) contention, this also indicates that a relatively more onerous regulatory environment does appear to make it more costly (and potentially prohibitively costly) for more marginal entrepreneurs to pursue growth enhancing opportunities that would otherwise be undertaken. Branstetter et al. (2014) find a similar result when evaluating Portuguese regulatory reform. Their results suggest that both firm formation and employment increase after reform, especially amongst marginal firms. Further, to the extent that these regulatory changes make pursuing such activities more costly, this also appears to translate into lower job creation emerging from those entrepreneurial endeavors that still do emerge. Finally, it appears that for those inframarginal entrepreneurs who still do pursue existing opportunities, while they do create fewer jobs at the margin, they're chances of surviving their first year of operation do not appear to be particularly dampened by the either Robust standard errors in parentheses. Constant suppressed from output tables ***p < 0.01, **p < 0.05, *p < 0.1 the current or accumulated regulatory burden. These are important results that could generate public policy insights and considerations in the future.

Conclusion
A large literature that evaluates how a jurisdiction's institutional environment impacts economic growth and development has evolved over time. This research tends to point out that it's the institutions or "rules of the game" that play a primary role in the economic outcomes that emerge across locales. Tied to this, is the importance that productive entrepreneurial activity plays and how fostering such activity is an important channel through which economies tend to grow over time. Thus, it is the existing institutional environment that can help or hinder such productive entrepreneurship and with it generate economic growth and development. The institutional features that exist are also directly tied to the regulatory environment within a jurisdiction. Here, where regulations exist to correct various market failures, then economic efficiency is promoted, which can and will ultimately enhance economic and entrepreneurial activity. However, where the regulatory environment becomes overly burdensome or onerous, this itself can generate negative consequences that may reduce growth in both the short-and long-run.
This current study adds to this latter body of research and evaluates how the federal regulatory environment within the US affects entrepreneurial activity and endeavors. Specifically, this paper applies a measure of the relative impact that federal regulation has across states and how this affects early-stage entrepreneurial endeavors and activities. Regulations applied at the federal level affect particular sectors across all states. However, an important consideration is to recognize that this impact can be different across states, depending on the extent to which a state's economy might be dependent upon a given industry. This can also then have differential impacts on and consequences for entrepreneurial activity across states as well.
With data from the FRASE index to measure federal regulatory burdens across states and several measures of early-stage entrepreneurship, this study finds that as federal regulations become relatively more burdensome within a state, there appears to be a reduction in opportunity entrepreneurship (i.e. pro-cyclical, growth-enhancing entrepreneurial activity), job creation, and early-stage entrepreneurship in general. However, there is no particularly strong evidence of a link between this regulatory burden and either the rate of entrepreneurship that emerges or early firm survival rates. These findings add an important piece to the literature, both confirming various previous studies and extending our understanding of the relationship between regulation and entrepreneurship in several important ways.
Author contributions John Dove is solely responsible for the completion of this manuscript.

Competing interests
The authors declare no competing interests.