People have produced such amazing products during the past few years that it is easy to forget that, after their unveiling, machines such as the printing press and the steam engine were viewed with just as much awe as the microprocessor is today. Yet what is clumsily referred to as "technological innovation" has consistently led to controversy as new products and processes replaced old ones-even in the 1400s. This dynamic process, central to the functioning of a market economy, was termed "creative destruction" by 20th century economist Joseph Schumpeter.

From the earliest days, the "destruction" portion of this dynamic process has brought into question government's proper role in a market economy. For instance, how far should policymakers go to "protect" the hand-copiers that the mechanical printing press would displace? To what extent should government implement policies to encourage entrepreneurs to invent the printing presses of tomorrow? These questions have been asked for centuries, but the answers still remain a source of debate.

Throughout U.S. history, most policy geared toward encouraging entrepreneurs has been passive. Instead of actively encouraging entrepreneurs with direct subsidies, U.S. policymakers have typically used government power to provide the laws and the institutional framework needed for individuals to run their own businesses. There is little doubt that much of America's economic well-being has resulted from this relationship.

For instance, Microsoft, General Electric, and Ford Motor Company were started by a handful of individuals with the government-guaranteed right to keep most of what they could earn from their innovations. Interestingly, these companies' successes have led many policymakers to pursue a more active role in encouraging entrepreneurial activity. Therefore, most current policy debates revolve around which active policies-e.g., targeted tax breaks, loan guarantees, and subsidies-work best. However, even tax policy can be viewed as a passive approach for encouraging entrepreneurial activity.

If all taxpayers pay the same income tax rate- regardless of the source of that income-then government policy is not giving preferential tax treatment to any one type of income. In this case, tax law passively encourages business activity by not providing tax incentives for (or against) starting a business. This paper provides a discussion of the research issues surrounding the impact that tax policies have on entrepreneurial activity.

Research Issues

On a theoretical level, judging how effectively tax policy influences business activity is not always straightforward. Measuring a successful outcome requires a clear definition of a policy's objectives. Simply stating that a policy's aim is to "encourage entrepreneurship does not provide an easily measurable objective because it relates to individuals' states of mind.[1] Even trying to measure such an objective using "increased employment," "increased startups," or "higher individual wealth" can be problematic.

Changes in employment are difficult enough to measure, but the newly self-employed present an interesting problem because, by definition, they do not work for another company. Consequently, even if a policy does cause individuals to leave their jobs and start their own companies, the net change in employment could-at least initially- be zero. Furthermore, when individuals start new businesses, their ultimate success or failure will likely remain unknown for several years.

Measuring success for those startup ventures that do last several years presents a problem similar to that of measuring employment. A measure of success for these long-lasting startups has to consider whether this business activity would have taken place in the absence of these particular companies. For instance, if Mr. Jones starts a new maid service company, a near-term measure of success has to consider any lost customers from Mr. Jones's competitors. Additional research difficulties, such as determining whether entrepreneurial activity increases the wealth of the business owners, are data-driven problems.

Data Sources

The most significant data problem in studying entrepreneurial activity is a lack of data. Although there are several data sources for studying demographic characteristics of small-business owners, the data needed to address most business-related tax policy issues are sparse. The following sections summarize the major data sources used to study business activity, and they provide brief descriptions of the strengths and weaknesses that each data set offers tax policy researchers.[2]

Statistics of Income. The Statistics of Income (SOI) division of the Internal Revenue Service (IRS) produces an annual Public Use Tax File that contains over 130,000 records. The SOI database is a statistically representative sample of all U.S. individual income tax returns for any given year. The data consists of virtually all major items that taxpayers provide on their returns, such as filing status, number of dependents, amount of deductions, amount of income from various sources (wages, dividends, etc.), adjusted gross income, and taxable income. Because the SOI file contains individual income tax data, the only business-related data in the file are what business owners report on their individual income tax returns.

For example, sole proprietors are required to file IRS form Schedule C; therefore, business income reported on Schedule C is in the SOI file. However, Schedule C contains roughly 50 line items and not all of the Schedule C information is included in the SOI file. For example, the SOI file does not include a separate total for repairs and maintenance expenses, but it does include a separate amount for wage expenses and for total expenses. Similar problems-and in some cases, more severe-exist regarding business information from other types of business entities, such as S corporations and limited liability companies (LLCs). In any event, the file does not contain important business information, such as the number of employees and cash on hand. Additionally, the annual releases cannot be used to track the same business owners across successive years.

For well-funded researchers, purchasing a longitudinal SOI file from the University of Michigan's Office of Tax Policy Research (OTPR) can mitigate this last problem. Michigan's OTPR has worked with the SOI to create a "panel" of tax return data that provides researchers with tax return data on the same taxpayers from 1979 to 1990. In addition to the public use file and the OTPR panel, the IRS posts on its Web site a limited amount of business-related data (taken from individual tax returns), as well as various aggregated tax information in the Statistics of Income Corporation Source Book.[3] However, because these data are already aggregated, they typically cannot be linked to individual taxpayers. Below are the strengths and weaknesses of the tax databases research:

• _div22_>Strengths: Provides a rich data source for information on U.S. individual income tax returns.
• _div22_>Weaknesses: Provides only limited business activity data.

  Panel Study of Entrepreneurial Dynamics. The Panel Study of Entrepreneurial Dynamics (PSED)[4] is coordinated by the University of Michigan's Institute for Social Research. The PSED is a longitudinal sample of nascent entrepreneurs and is uniquely designed. The "panel" consists of 830 individuals actively engaged in starting new businesses. Importantly, it tracks these same individuals for two years following their startups. In other words, the PSED contains three years of data on new entrepreneurs, and the first year of data covers the first year that the business existed.

  The PSED is the first U.S. database to offer longitudinal data that can be used to describe the characteristics of individuals engaged in the process of starting a business. Most of the data in the PSED relates to social characteristics and demographics. For example, the nascent entrepreneurs report age, gender, ethnicity, education level, family composition, income, and net worth. Additionally, the PSED contains data that may explain personal preferences for starting a business, such as family business background, work history, and business climate perceptions. The PSED does collect some data directly concerning policy issues, such as whether or not (and what type of) public assistance programs were used to start the business venture, but the bulk of the data consists of social and demographic characteristics.

  Because the PSED examines startup ventures, most financial data relate to the owners' projections rather than actual values. For instance, many startups do not have any sales (or even a marketable product) for the first few years of existence. Similarly, although the PSED does contain information on initial funding sources, judging the need for additional funds and/or financial assistance initially depends on the owners' projections. Finally, the PSED is not designed to study specific tax policies and, as such, does not ask nascent entrepreneurs to report any detailed tax information. Below are the strengths and weaknesses of the PSED research:
• _div22_>Strengths: Offers a wide array of social and demographic characteristics for individuals engaged in starting business.
• _div22_>Weaknesses: Contains very limited tax-related data; does not collect a robust set of business data from going-concern entity owners; and includes a relatively small sample size.

  Survey of Business Owners and Self-Employed Persons. The U.S. Census Bureau conducts the Survey of Business Owners and Self-Employed Persons (SBO) every five years, and the latest survey contains data from 2002. The SBO contains business owners' demographic information, such as the gender, race, age, ethnicity, and educational level, as well as economic data such as primary business function, types of customers, types of workers, source of startup capital, and source of financing for capital improvements.

  However, the SBO questions are rather general in nature, and business-owner data are not released to the public. For example, the survey does not ask for the amount of startup capital or the dollar amount of capital improvements made in a given year. Similarly, the SBO asks whether workers are full-time or part-time employees, but it does not ask how many are employed. Finally, the SBO does not contain any income or tax-related data. This survey was previously conducted as the 1997 Economic Census Surveys of Minority-Owned Business Enterprises and Women-Owned Business Enterprises.[5] Below are the strengths and weaknesses of the SBO research:
• _div22_>Strengths: Provides a wide range of social and demographic characteristics of business owners.
• _div22_>Weaknesses: Data are collected only every five years, and only aggregate-level data are released to the public.

  Longitudinal Business Database. The Longitudinal Business Database (LBD)[6] is a dataset constructed by the Bureau of the Census' Center for Economic Studies. It is not publicly available. The LBD is a longitudinal database that follows individual businesses (with paid employees) from 1975 to 1999. The LBD is principally used to study trends in industry entry and exit, gross job flows, and overall changes in the structure of the U.S. economy.

  These data are complied from a variety of sources, such as the federal Business Register, the Economic Censuses, and various surveys. Information in the LBD is collected at the establishment level-meaning that companies with multiple locations are treated as multiple establishments. Even though the file attempts to "link" establishments over roughly 25 years, most of the establishments in the LBD remain in the data for only a few years.

  As is the case with most business-related Census Bureau surveys, the LBD tends to gather general information on payroll, employment, and sales, but does not gather detailed tax and income information on individual business owners. Below are the strengths and weaknesses of the LBD research:
• _div22_>Strengths: Contains business data for the same business establishments across multiple years.
• _div22_>Weaknesses: Does not include detailed tax and income information about individual business owners.

  Current Population Survey. The Bureau of the Census conducts the Current Population Survey (CPS)[7] for the Bureau of Labor Statistics (BLS). The CPS surveys about 50,000 households each month, and it is the primary source of information on labor force characteristics for the U.S. population. (The CPS data are used to produce the monthly BLS Employment Situation report.) The sample provides estimates for the nation as a whole and serves as part of model-based estimates for labor force characteristics in individual states and other geographic areas.

  The CPS collects data on income, employment, unemployment, earnings, hours of work, and a variety of demographic characteristics including age, sex, race, marital status, family structure, and educational attainment. Information reported in the CPS also allows researchers to sort data by occupation and industry, and supplemental questions are frequently added to study timely topics, such as employee benefits and health plans. However, the CPS does not collect detailed business information, such as sales, capital expenditures, and asset size. Additionally, specific tax data items-such as Schedule C wage expenses and taxable income-are typically not included in the CPS. Below are the strengths and weaknesses of the CPS research:
• _div22_>Strengths: Contains a wealth of income and demographic information for a nationally representative sample of U.S. households.
• _div22_>Weaknesses: Does not gather detailed information about business establishments and does not provide a large sample of business owners.

  Longitudinal Employer-Household Dynamics. The Longitudinal Employer-Household Dynamics (LEHD), also know as the Local Employment Dynamics,[8] is conducted by the Census Bureau. The LEHD survey collects data on increases and decreases in employment on a quarterly basis in 10 states.[9] The LEHD is compiled using state unemployment insurance wage records and the Department of Labor's ES202 data.[10]

  The LEHD's main goal is to provide its partner states with improved data on changes in workforce composition. The establishment-level LEHD data are not publicly available, and detailed business information such as sales revenue, profit, capital expenditures, and funding sources are not collected. Below are the strengths and weaknesses of the LEHD research:
• _div22_>Strengths: Provides employment data for the same business establishments across multiple years.
• _div22_>Weaknesses: The survey is not conducted nationally and the results are not publicly available.

  VentureXpert. Thomson Financial's VentureXpert database is a main data source for researching the private equity industry. For example, the database contains information on initial public offerings (IPOs), venture capital deals, buyouts, and limited partnerships around the world. Most of the data included in VentureXpert are time-series data, such as the number and sizes of deals completed for a given month and the number and share prices of IPOs in a given quarter. Below are the strengths and weaknesses of the VentureXpert research:
• _div22_>Strengths: Provides a fair amount of data on private financial market transactions.
• _div22_>Weaknesses: Because it is primarily a financial market database, it contains virtually no information about business owners and does not track business activity.

  Public Financial Records. Publicly traded corporations are required to file annual financial reports with the Securities and Exchange Commission, and these reports are a matter of public record. These reports contain a wealth of financial accounting data-such as asset size, sales volume, debt, and cash flow-but they do not contain information from these companies' tax returns. (Tax returns are considered private information.) Most publicly traded corporations are owned by more than just a few individuals, and most are quite large in terms of assets, sales volume, and/or number of employees.

  Data collection from these reports can be quite tedious because each company's report contains only information for that particular company. Furthermore, public companies report financial information based on Generally Accepted Accounting Principles. Because not all businesses operate in an identical manner, these rules offer companies leeway with many reporting choices. Some of these problems can be overcome by purchasing Standard and Poor's Research Insight database; a product that "standardizes" financial statement data from virtually all publicly traded companies. Additionally, corporate income tax simulation models can be created by combining data from the Research Insight database with tax data from the Statistics of Income Corporation Source Book. These models allow researchers to study corporate income tax code changes in a manner similar to the way that they study individual income tax policies. Below are the strengths and weaknesses of public financial records:
• _div22_>Strengths: Contains an enormous amount of financial accounting data.
• _div22_>Weaknesses: Contains very little corporate tax return data and virtually no social, demographic, or income data on business owners.

  The Leading Policy Issues

  These datasets provide limited, though important, opportunities for research about the relationship between entrepreneurial activity and federal tax policy. Among these opportunities are analyses of the tax policy effects on how entrepreneurial businesses are organized, the role that tax policy plays in changing the capital costs that entrepreneurs face, and how tax policy shapes the savings behavior and choices of entrepreneurs.

  One of the best-known effects of tax policy is the influence that it exerts in the choice of organizational form. Entrepreneurs who require the legal protections and large-scale operating advantages of a corporate business organization can choose from a variety of corporate organizations. As shown in the following section, entrepreneurs and business managers have exercised these choices as tax rate differences emerge between different forms of organization.

  The level and type of entrepreneurial activity is often affected by the cost of capital. That is, indirect and direct taxes on capital can raise the cost of capital to borrowers above the economic cost that lenders require to compensate them for the temporary loss of the funds and the risk of investment. These higher costs can raise substantial barriers to startups and expanding businesses that lack the cash flows or credit ratings required to borrow expensive capital.

  Differentials in capital costs that stem from differences in tax rates can also produce challenges for entrepreneurs. Falling tax rates are always welcome news to entrepreneurs. However, capital goods (e.g., drill presses, computers, and transportation equipment) placed in production with highly taxed capital take longer to pay off than the same goods taxed at lower rates. Thus, older manufacturing firms are generally less profitable than newer ones from the standpoint of the after-tax productivity of their equipment.

  Just as differences in tax rates can distort the value of otherwise identical equipment, they can also heavily influence the life and future of existing firms. Estate or death taxes have a particularly striking influence on entry and exit decisions. For example, small business entrepreneurship is the common choice of new immigrants, minorities, women reentering the workforce, and retired people needing extra income. In each case, success could mean paying estate taxes that approach 50 percent. Indeed, women, immigrants, and minority business owners have frequently cited federal estate taxes as a major factor in their decision to expand and continue.[11]

  There are doubtlessly many challenges that entrepreneurs face because of the current set of federal tax policies. However, those described above offer promising opportunities for breakthrough research in the policy research field of the interplay between taxes and entrepreneurship.

  Picking Winners and Losers Is Counterproductive

  Many policymakers favor "targeting" tax relief to certain types of businesses to promote growth in what they view as "key" industries. Of course, this sort of policy begs at least two questions: Why not help other types of businesses? Which types of businesses should be helped? Favoring any particular industry over another is the wrong approach because it makes reforming the tax code more difficult, distorts market incentives, and leads to an inefficient allocation of resources.

  For example, when business owners make decisions based solely on tax credits, they spend money on goods that they otherwise would not have purchased. These purchases may appear to boost the economy, but they have hidden consequences. If a technology tax credit causes business owner A to buy a new computer from business owner B, business owner B will happily make the new sale, but what about the owner of a non-technology-related store?

  The net economic gain from any money spent on new computers will be at least partially offset by the money not spent on other goods. This hidden impact is too easily forgotten, but it is no less real than the obvious impact from improved incentives in the computer market. Additionally, these types of targeted incentives make reforming the tax code more difficult because they create constituencies that will lobby to keep special tax advantages.

  For example, the recently passed American Jobs Creation Act of 2004 (H.R. 4520) includes at least eight separate provisions dealing with depreciation rules. One of these provisions extends "bonus" depreciation eligibility by one year for certain non-commercial aircraft put in service before January 1, 2006. Manufacturers of these aircraft are therefore likely to resist any tax reform that proposes to eliminate all targeted depreciation rules. Because so many special provisions already exist within the tax code, even small steps toward simplifying the code pits winners against losers-making reform that much more difficult.

  Regardless of the exact nature of a reform plan, researchers should evaluate the plan by quantifying the expected value of businesses' future tax benefits-a task that is not easy. Because so many competing interests are invested in special tax breaks, removing all business taxes may be the reform policy that best minimizes this problem. Various tax constituencies would lose some special benefits under such a reform plan, but the long-term need for those benefits would be eliminated. Of course, there would be serious political obstacles to any reform plan that eliminates all business taxes.

  There Is No "Best" Definition of What Constitutes a Business

  One of the many squabbles in the last election cycle centered on whether to extend or to rescind President George W. Bush's tax cuts for taxpayers paying the top marginal tax rate. An argument against rescinding these cuts was that many small business owners would be hurt because they pay individual income taxes at the top marginal tax rate. Curiously, part of this policy debate centered on which business owners were operating "real" businesses.

  Business income from a wide variety of business types can show up on individual income tax returns. For instance, a group of doctors could organize a partnership to buy a beach house in Florida. If they periodically rent the house to tourists, the doctors will end up with partnership income on their individual tax returns. On the other hand, someone might organize a construction company as an S corporation and be actively involved in performing physical labor. Advocates of rescinding recent tax cuts argue that only the second of these two examples should be considered a business.

  Yet why are both examples not legitimate businesses? In both cases, individuals have organized to better invest their money and use their resources. Both businesses will buy products and, either directly or indirectly, employ other people. In fact, the construction company may even perform services for the partnership. Each of these individuals is undertaking legitimate productive activities that benefit others. Disparaging either legal entity as "not a real business" ignores this fact. The debate about the "real" business owners also took on a class warfare theme, with tax cut opponents arguing for raising taxes on "wealthy" business owners.

  

  

  Income Is a Fleeting Concept in the IRS Data

  Supporters and critics of the Bush Administration's tax policy label taxpayers earning more than $200,000 as "the wealthy," as if this were an easily identifiable group of taxpayers. This notion is problematic because the definition of income is not straightforward and because identifying taxpayers in any given year-for any purpose- ignores long-term trends.

  To begin, tax policy researchers typically define income as adjusted gross income (AGI) because it provides a common point of comparison. However, AGI is a special IRS definition of income, and it accounts for up to 18 types of income. Just as important, AGI is calculated after any number of deductions: Form 1040 lists 10 deductions, and all business income from pass-through entities is reported after business expenses have been deducted. Even granting that AGI is the best measure of income, choosing a level of $200,000 as the cutoff point for tax breaks is completely arbitrary and is no better or worse than $175,000 or $225,000.

  Furthermore, IRS data are reported on an annual basis, obscuring information that could be learned by following taxpayers' income over a number of years. For instance, the number of business owners who typically earn AGI above $200,000 could be understated in any given year simply because of unusually high business expenses. Even conceding that the number of taxpayers with more than $200,000 in AGI for any given year is the true number of such taxpayers, the number of taxpayers in this category can change in future years simply because a new law eliminates certain deductions.

  If multiple legal changes are made during several years, the number of taxpayers in a given AGI class could change for reasons unrelated to economic activity. Regardless of how "income" and "business" are defined, tax policy that singles out any group for higher taxes hurts all taxpayers. This rule holds true for owners of both large and small businesses, even though large corporations are frequently vilified for political reasons.

  

  Large Businesses Were Not Always Large

  Treating owners of both small and large businesses the same will certainly be criticized as "pandering to corporate America," and ignoring the "little guy." However, championing small business owners while simultaneously disparaging owners of large businesses is contradictory. Owners of large businesses start out by owning small businesses. People are not born owning businesses: They choose to start companies at some point in their lives based on their experiences.

  Tax policies that intrude on the decision to start a business-or on decisions made while running a business of any size-distort decisions that would be made in the absence of these policies. Therefore, an economically neutral tax policy is one that does not discourage potential business owners, small business owners, or large business owners from making economic decisions. In other words, the best tax policy affects all business owners the same.

  Taxing income only at the individual level would be a significant departure from the current tax code, but the idea is based on a simple premise: Corporations do not pay taxes-people do. Corporations of all sizes are merely legal entities. They are all run by people and they all sell goods and services to people. Ultimately, all corporate taxes are taken out of the pockets of people, either through lower compensation to workers or higher prices paid by consumers. Historical data on new-business formation suggests that more and more business owners understand this concept quite well.

  

  Most New Business Owners Organize As Non-Corporate

  More and more business owners are choosing to organize as "pass-through" entities, such as S corporations and limited liability companies. These "non-corporate" forms afford legal protection similar to that of traditional corporations, but they allow business income to pass through to the owners' personal tax returns. Consequently, pass-through entity owners' business income is taxed only at the individual level, whereas owners' income from a traditional C corporation is taxed at both the corporate and personal level. Historical data show increasing trends in the number of non-corporate entities, and the owners of these businesses continue to pay taxes on their business income-just not through the corporate tax system.

  Chart 1 shows the trend in the number of S and C corporation filings from 1975 through 2004.[12] In 1975, S corporations accounted for only 17.22 percent of the total returns, while C corporations accounted for more than 80 percent. However, by 1996, S corporation returns accounted for the majority of the total. The IRS projects that S corporation returns will account for nearly 60 percent of these business tax returns in 2004.[13] Table 1 demonstrates that there has been substantial growth in filings for other types of pass-through entities as well.

  Both sole-proprietor and partnership (including LLC) filings have more than doubled from 1975 to 2004. During this same time period, traditional C corporation filings increased by only 23 percent. Table 1 also provides a more recent comparison that shows that the number of C corporation filings actually declined from 1990 to 2004. In contrast, the growth in all three categories of pass-through entity filings increased substantially from 1990 to 2004.

  The number of sole proprietors-defined as individuals filing an IRS Schedule C, C-EZ, or F- increased by 28 percent from 1990 to 2004, while partnership returns increased almost 42 percent.[14] During this same time period, the number of S corporation filings increased 127 percent, while C corporation filings decreased 7 percent. Given the enormous relative growth in the pass-through entities since 1990, the distinction between corporate and individual taxes is much less meaningful than it was only 25 years ago. Consequently, business-related tax policies have to account for provisions in the individual income tax code. Statistics about the level of business activity that these non-corporate entities contribute to U.S. economy further support this notion.

  Statistics About "Non-Corporate" Businesses

  There is no perfect measure of how important any group of businesses is to the U.S. economy, but there are some basic figures that provide insight into this inquiry. Table 2 summarizes the total amount of wages paid by pass-through entity businesses in 2001. These data show that S corporations, partnerships, and sole proprietorships paid out roughly $665 billion in wages.[15] This figure probably understates total wages paid out by these entities, because not all S corporations and partnerships are required to submit financial statements to the IRS. Nonetheless, this $665 billion in wages represents almost 20 percent of all private wages in the U.S.[16]

  Of course, wages are not the only measure of how important businesses are to the U.S. economy. Charts 2, 3, and 4 demonstrate that owners of pass-through entities operate in a wide array of industries, ranging from real estate and health care to construction and financial services. Just as important, the financial records of each entity group show that these businesses rely on multiple sources of income.

  Across industries, these businesses derive significant amounts of income from sources such as financial market investments and investments in other businesses. (See Tables 3 and 4.) These relationships suggest that over-burdening any particular industry with higher taxes (relative to other industries) would be felt by all types of business owners and their customers. A similar case can be made for owners and customers of corporate entities.

  

  

  

  

  

  Statistics About "Corporate" Entities

  A majority of the economic activity in the U.S. takes place through publicly traded corporations. However, as with non-corporate businesses, there is no perfect measure of this activity. Table 5 summarizes several financial accounting measures taken from Standard and Poor's Research Insight database. These figures are from 2002 and were reported by all companies (in the database) with total assets greater than $175 million.[17]

  This group of corporations reported total (book value) assets of about $46 trillion, from which approximately $14.6 trillion in net sales was generated. These companies reported $933 billion in capital expenditures; more than $2 trillion in selling, general, and administrative (SG&A) expenses; and $9.9 trillion in "cost of goods sold."

  Publicly disclosed financial statements also provide measures of income generated for shareholders. For example, the amount of cash dividends these companies paid to shareholders in 2002 was more than $304 billion.[18]

  Another measure worth examining is free cash flow, a measure that represents the cash left over after all dividends, expenses, and capital expenditures are deducted. These firms reported a free cash flow of almost $520 billion in 2002. Although this number is quite large, it represents only 4.35 percent of the companies' combined cost of goods sold and SG&A.

  These figures do not support the notion that large corporations have huge economic profits and can therefore easily absorb tax increases. Furthermore, neither the size of these companies nor their form of legal organization changes the fact that they are only legal entities. These organizations, just as the pass-through entities discussed above, are run by people and sell goods and services to people. Taxing either type of entity ultimately places an additional financial burden on people. (See Table 5.)

  Conclusion

  Throughout most of U.S. history, governments at all levels have primarily strived to establish a rich and stable institutional setting for entrepreneurship numerous and open courts for adjudicating legal problems, a banking system with safeguards and flexibility, tariff-free movement of goods and labor across state borders, and so forth. Although major exceptions exist, the grand drift of American history places the individual as the actor in a marketplace that states protect but do not shape financially.

  That relationship between government and entrepreneurs fundamentally changed in the 20th century, particularly after World War II. The federal and state governments used many policy handles in their retreat from laissez faire economic policies, but the federal tax code stands out as the most powerful of all of the tools. Markets, costs, economic incentives, and even whole categories of goods and services were changed by the way the federal government taxed enterprises.

  Given that influence, it is surprising how little attention has been paid to the relationship between entrepreneurship (at all levels) and federal tax policy. As this paper indicates, numerous databases could-and do-support research into this relationship, and the policy issues they could inform touch on some of the most important topics in tax economics. Research on these issues promises to produce insights that could well shape tax policies in order to strengthen entrepreneurship and economic activity.

  Norbert J. Michel, Ph.D., is a Policy Analyst in Macroeconomics and Ralph A. Rector, Ph.D., is a Research Fellow and Project Manager in the Center for Data Analysis at The Heritage Foundation.