Today San Francisco faces the dual curse of budget woes and economic stagnation. An unfortunate result of the combined impact of national recession, the dot-com crash, and increased public spending, San Francisco 's budget trouble will not be easily solved.
Raising business taxes may be one way to close the gap. But raising business taxes is also risky: raising them too high, or in the wrong way, can potentially drive jobs from the city and hurt the already shriveled tax base
Calibrating the business tax in the right way, balancing the goal of revenue generation with the goal of not discouraging job creation, is of critical importance for the fiscal well-being of the city. Broadly speaking, the major issues related to business taxes in San Francisco are:
• A bad economy. San Francisco 's economy is troubled. At the end of 2003 commercial vacancy was at 20.37%-16.89 million square feet-just off from the city's all-time high of 20.6% and nearly the highest in the nation (BT Commercial Real Estate, courtesy SF Chamber of Commerce). Although traditionally a strong spot, now even San Francisco 's tourist industry is faltering, as hotel vacancy rates have also hit all-time high levels. The city has lost more than 60,000 jobs, giving back nearly all the jobs created in the dot-com boom. It is too soon to tell if the decline is temporary and cyclical or if it portends more serious underlying problems with the city's economy.
• The legal challenge. From 1970 until Spring 2001 San Francisco maintained an alternative-measure business tax, consisting of a 1.5% payroll tax and a varying rate gross receipts tax. Every company doing business within the city calculated its tax payment under each system and paid whichever amount was higher.
But in 1999, General Motors and Eastman Kodak brought suit against Los Angeles , which had a similar alternative-measure tax, claiming it was unconstitutional; in March 2000 the California State Courts concurred. In essence, it was decided that levying the gross receipts and payroll taxes against different firms simultaneously was unconstitutional. One or the other would be allowed, but it is illegal to do both.
Some of the same companies sued San Francisco , and the city decided to settle. The San Francisco Board of Supervisors passed a measure to eliminate the gross receipts tax, paid by 18% of the city's businesses, effective May 25, 2001 . Subsequent efforts to more carefully revamp the business tax system-culminating in Proposition I on the November 2000 ballot, which would have enacted a retroactive payroll tax-have failed.
The costs of this stopgap solution have been great. The gross receipts tax repeal resulted in a net revenue loss of approximately $20 million annually, including $16.5 million in tax payments and $3.5 million in registration fees. In addition, the city will have to pay $7 million in debt service on bonds issues to pay settlement costs until 2011. The city has paid approximately $60 million in settlements related to the lawsuit to date.
• Spending. San Francisco 's City budget grew by 70% between 1995 and 2003, a rate three times faster than inflation. Increased spending was largely the result of transportation and other infrastructure improvements, as well as continued spending on public health (King 2001) . During this period, Muni service was improved to make up for past cutbacks, major modernization of the city's water system was undertaken, and the city's airport was expanded. While these expenditures may create positive benefits for San Francisco companies, their impact on the city's budget is substantial.
As of this writing, San Francisco 's projected budget deficit for fiscal year 2004-2005 is $260 million. These problems are magnified by the ongoing political debate on the optimal size of the public sector in San Francisco . Some argue that a high-tax, high-service model, along European lines, is best for the city, while others advocate a lower-cost, free-market environment.
The contribution of business taxes to the overall City budget is substantial: business taxes constitute the fourth largest source of revenue to the General Fund and the sixth largest source to the overall budget.
The considerable weight of business taxes in city revenues means that changes in levels of business taxation can have a large impact on the overall city budget.
For all of these reasons, the matter of business tax policy has come to the fore. This report aims to inform the debate on this critical issue by answering three basic questions:
• What does the academic literature say on the impact of business taxes on firm location?
• What kinds of business taxes are the "best" taxes?
• How do San Francisco 's business taxes compare to those levied in comparable cities?
These questions define the organizational structure of this report.
This report is a literature review, meaning it is not original research. Instead, it digests, summarizes, and analyzes a wide range of academic work. Where appropriate and necessary, secondary sources are noted using Chicago-style in-text citations matched by full citations in the comprehensive bibliography located at the end of the report. Boxes periodically interrupt the text; these define concepts and terms that may be unfamiliar to some readers.
I. Business Taxes and Firm Location
More than ever before, corporations and capital are mobile. Economic deregulation, communication and transportation technologies, and increasing international trade have allowed companies to locate , relocate , and expand wherever they expect to find the highest return on their investment. Increasing "globalization" and the consequent mandate that firms be competitive in a global environment have essentially required that for-profit companies, which are seeking to attract investment of capital, treat locational decisions as an explicit part of their business strategy.
What is Globalization?
The term "globalization" describes the increased mobility of goods, services, labor, technology and capital throughout the world. Although globalization is not a new development, its pace has increased with the advent of new technologies.
These changes mean that a firm's location is now a factor that is continuously reevaluated. The implications of this change are profound for governmental jurisdictions. Most important are competitive forces that have prompted a spatial reorganization of economic activity:
• Cities are on a global stage competing for jobs and capital. With international, national, and regional competition, the overall "marketplace" for firm locations is very tight.
• There has been increasing diversification by function. That is, different activities of the firm are being moved to where costs for that particular function are lowest. For instance, manufacturing has been moved to the developing world and "back office" administrative functions (e.g. call centers, payroll, and billing) have been moved to attractive but low cost areas such as Arizona and Utah, and increasingly, English-speaking countries with high levels of education such as India. Diversification by function means that cities must worry not only about keeping or attracting a company as a whole, but also about keeping or attracting diverse, separately mobile parts of that firm.
• The pressure on cities to contain tax costs is enormous, and as a result, competition in the form of tax breaks and other economic incentives has become fierce. For every medium- or large-sized firm unsatisfied with its present location, there are a handful of jurisdictions willing to offer it deep tax discounts in order to relocate there.
Components of the Location Decision
It is important to remember that tax costs are not the only factors influencing firm location. In fact, a host of issues-really anything that impacts a company's costs , revenues , or productivit-may be important. Broadly speaking, these factors are:
• Prevailing wages. There are great differentials in wage levels between certain job markets. Firms seek markets where they can employ the type of workers they need at the lowest possible wages.
• Labor force quality. There are also great differentials in levels of skill and education between cities and regions. Firms seek regions where workers with the necessary skills and education are available for the right cost.
• Housing and quality of life. Related to wage rates and labor force availability are concerns over housing prices and general quality-of-life factors. Firms want to locate where their employees can afford good housing and where their quality-of-life demands can be met. This may include the quality of schools, healthcare, recreational opportunities, and climate.
• Labor market rigidities. Restrictions on the ability of managers to allocate labor, generally associated with levels of unionization, vary especially between older regions and newly developing ones. Firms prefer to avoid doing business in places where unions are powerful, especially when less-unionized alternatives are available.
• Proximity to suppliers or final market. The location and accessibility of firms supplying inputs and tools as well as the location and accessibility of markets for finished products are of great importance to some firms. This may not only mean physical proximity, but also access to ports, airports, and other transportation infrastructure.
• Energy and resource costs. The cost of electricity and fossil fuels, as well as the availability of water and other resources, varies greatly between regions. Firms for which these costs are large seek locations where the energy and other natural resources they require are available at low rates.
• Real estate costs. All firms require office space and many require industrial or other operational space. Space costs, which depend on overall conditions in a region's real estate market, vary between regions. Firms desire high-quality space at reasonable prices.
• Political stability. In some places, political conditions mean that laws change frequently and that there is a high turnover of elected officials; in other places there is greater continuity. Firms prefer to make decisions in an environment of stability and predictability.
• Innovation capacity. Some regions have a greater capacity for "innovation" than others. Innovation is a loosely applied concept used to describe the special abilities of firms, labor, and regions to implement value-creating technological, financial, and organizational ideas. Affecting regional innovation are factors such as education levels, social characteristics, and industry's relationship to universities. Some firms prefer to locate in regions they see as innovative.
What is economic innovation ?
Innovation is the conversion of knowledge and ideas into a benefit, which may be for commercial use or for the public good; the benefit may be new or improved products, processes, or services. Innovation and technological change are main drivers of economic growth at organizational, sector, and economy-wide levels.
• Agglomeration benefits. Agglomeration benefits are efficiencies realized by firms who "cluster" with other firms and consequently obtain better access to supply inputs, a labor force, and markets, as well as other more dynamic advantages such as innovative capacity (Porter 1996) . Scholars recognize two types of agglomeration benefits: those that accrue to firms by locating near other firms in the same industry ("industrialization economies"), and those that accrue to firms by locating near firms in other industries ("urbanization economies").
Finally, it is the goal of this report to digest the academic literature on how and to what extent local taxes influence firm location:
• Tax costs. Locally variable taxes affecting businesses include corporate income (and profit) tax, property tax, taxes on business activity such as payroll and gross receipts tax, utility taxes, business licensing fees, sales taxes, fees to special districts (e.g. the East Bay Municipal Utility District), and so on. Taxes levied on individuals may also affect the firm-location decision, as they affect the firm's employees.
• Public services. As the other side of the tax equation, public services include a wide range of activities that directly and indirectly support business. Direct benefits of public spending by local government include police and fire protection, infrastructure provision (such as sewers), transportation (public transit and roads), business development support, and sanitation. Indirectly beneficial services include public education, land-use planning, and even public assistance. Public services attract economic activity because they either provide an un-priced input to production or, although a service is not directly used by business, greater quantities of the public service are associated with a lower price for some input that is used by business (Bartik 1991) . For example, good parks or other quality-of- life amenities can be helpful in attracting employees.
Because taxes pay for something-services-it is too simple to say that a firm simply seeks the lowest tax costs. Certainly a firm would want to avoid a high-tax jurisdiction in which tax money was inefficiently handled or spent on programs that are meaningless to businesses, but beyond this it is difficult to say whether a particular firm would prefer a high tax/high service environment or a low tax/low service one. For instance, Bartik (1999) notes that many studies have found that a budget that balances an increase in taxes with an increase in spending on education and roads will actually increase economic activity. In general terms, a policy of tax increases/increased spending or tax cuts/decreased services would have indeterminate effects. For these reasons, taxes and public services are best viewed as a single, intertwined issue (Bartik 1999).
Looking at the Literature: The Three Major Subgroups
Many studies have been published in the ongoing academic debate on the impact of taxes on firm location and economic growth. This section reviews the key studies, providing an overview of the important works.
Because the literature is vast, certain criteria were applied in choosing the most relevant works. All the literature reviewed here:
• Address locally variable taxes. Some studies focus on the impact of tax differentials between nations or U.S. states, but our concern is more local. This literature review was limited to studies that focus on types of taxes that vary at the county level or, preferably, the city level.
• Address firm location or growth. There are many different measures of business activity and economic growth through which to study the impact of taxes. These measures include income, office space absorption, new business permits, and so on. While studies included in this review use a wide variety of measures, those that use job creation and other more direct indications of firm location are the focus.
• Are relatively recent. Because methods of analysis and overall economic conditions have changed rapidly in recent years, the focus in this literature review will be studies conducted since around 1990.
This literature review divides the studies into three major "categories":
• Studies that seek tax/growth elasticities , which look for a numerical relationship between tax rates and growth.
• Recently, a group of researchers have sought to define local revenue hills (revenue curves), asserting that after a certain point, raising taxes discourages economic activity enough to actually decrease a jurisdiction's tax revenue.
• Finally, there are a number of studies and articles that look at real world firm-location decisions , trying to assess the circumstances around specific location decisions and the location decision-making process in general.
It is important to note that most studies distinguish between inter-metropolitan and intra-metropolitan firm location. Generally, studies find that on the inter-metropolitan level-when companies are deciding between locating in different regions (e.g. the Bay Area vs. the Seattle area)-taxes are less important than when a firm is choosing between different municipalities within the same region (e.g. South San Francisco vs. San Mateo). This is because many of the non-tax factors in firm location, such as wages and workforce skills, are essentially constant among cities in the same region. Tax rates, on the other hand, still vary greatly. The inter- vs. intra-regional distinction is a recurring theme in all the literature and will be explored in each section. And as Section III (Tax Comparisons) of this report will show, San Francisco has a hard time competing on taxes alone with other Bay Area jurisdictions.
Tax/Growth Elasticity Studies
The majority of studies that examine the impact of taxes on firm location are highly quantitative and econometric.
What is econometric study?
Econometrics is the application of mathematical and statistical techniques to economics in the study of problems, the analysis of data, and the development and testing of theories and models.
For the purposes of this report, we use Mark, McGuire, and Papke's narrower definition: An econometric study is one which-relates a variable of interest (a dependant variable), such as branch plant openings or aggregate employment growth, to several variables (the independent variable), such as electricity costs or the quality of the labor force, that are theoretically expected to influence the variable of interest. This method of analysis enables the researcher to determine, in a rigorous way, which of the independent variables are statistically significant factors for explaining the dependent variable- (Mark, McGuire, and Papke 1998) .
Generally speaking, econometric studies that examine growth and location impacts of tax differentials seek to determine the tax elasticity of economic activity (the response of one attribute of economic activity to another's rise or fall with respect to tax rates).
What is elasticity?
Elasticity is a concept in economics that measures the responsiveness of one variable in response to another variable. The best measure of this responsiveness is the proportional or percent change in the variables. Thus elasticity is the proportional (or percent) change in one variable relative to the proportional change in another variable.
The general formula for elasticity is: E = percent change in x / percent change in y
It is important to note that each study measures a very specific attribute of economic activity (e.g. creation of industrial space, absorption rates of office real estate, net job creation, investment) as an indicator of growth and economic activity in general. This is one important reason why caution is due when comparing absolute findings of elasticity.
These elasticity figures indicate how much changes in tax rates influence economic activity. Where high elasticities are found, taxes have a strong impact on growth and firm location; where low elasticities are found, taxes have a weak impact. Where elasticities are very low, there may be no impact at all.
When an economist finds a very low elasticity, how does she know that the elasticity is not effectively zero? To answer this question, economists apply the statistical tool of significance . Significance testing indicates whether a result is likely to be the product of mere random sampling.
In the context of these studies, significance testing helps indicate if taxes have an impact on economic activity at all. When a result is found to be "significant," we say that taxes matter; when a result is "not significant," we say that taxes do not matter.
What is statistical significance ?
A finding is described as statistically significant when it can be demonstrated that the probability of obtaining such a result by chance only is relatively low. In economics it is customary to describe one's finding as statistically significant when the obtained result is among those that would theoretically occur no more than 5 (or sometimes 1) out of 100 times when the only factors operating are the chance variations that occur whenever random samples are drawn.
Mark, McGuire, and Papke add an important qualification: -Statistical significance should not be confused with the size or importance of the effect. A variable can be statistically significant, and thus-matter,' but the estimated coefficient and elasticity may be small, in which case we would say that the variable is not an important factor- (Mark, McGuire, and Papke 1998) .
The range of elasticities that have been found for tax impacts on firm location is wide. For example, Grieson (1980), Gruenstein (1980), and Inman (1987) tested the employment effect of the Philadelphia wage tax very similarly. Gruenstein finds a very small effect on total employment-a wage tax elasticity of employment between -.01 and -.04 (total wages paid decrease by between 0.1% and 0.4% for a 10% increase in taxes). Inman, on the other hand, finds a substantially greater elasticity-between -.11 and -.14 (1.1%, 1.4% decrease per 10% tax increase). Grieson, who disaggregates the industrial sectors, finds the largest elasticities-between -.23 and -2.1 (2.3%, 2.1% decrease per 10% tax increase).
The wide range of these elasticities may indicate that not all studies are conducted using the same variables and robustness, or that taxes behave differently in different places and on different types of firms. In practice, the cause of the wide range is probably a mix of both these factors.
Major Elasticity Literature Reviews
More important and reliable than looking at specific elasticity figures from disparate studies are characterizations of different studies' findings as indicating that taxes matter or do not matter, and interpretations of how much they matter. Additionally, studies that look at a wide selection of studies' elasticity findings may suggest a reasonable "ballpark" elasticity figure. A number of literature reviews that look at econometric studies on tax impacts on growth have already been conducted; the most important such reviews are described here.
• Bartik (1991). Based on 48 studies, if a state or metropolitan area reduces state and local business taxes by 10% without changing its public services, and without changing their fiscal policies, then business activity in that state or metropolitan area on average seems to increase in the long run by around 3%. In other words, the long-run elasticity of a state or metropolitan area economic activity, with respect to state and local business taxes, is on average estimated to be around -0.3.
• Bartik includes a chart summarizing the elasticity findings of various studies and his statistical analysis these findings. Following is a simplified version of this chart:
Bartik's Summary of Econometric Studies of Tax Effects on Business Location
# of studies
Percentage of studies with at least one statistically significant negative tax effect
Mean elasticity of business activity with respect to taxes
Range of elasticities of business activity with respect to taxes
-1.40 to -.76
Inter-area studies with controls for public services
-1.40 to -.76
-4.43 to -.62
Intra-area studies with specific community data
-4.43 to -.62
From this chart we can see that the impact of tax increases on economic activity is much greater on the intra-regional level than on the inter-regional level . That is, competition is far more stiff between San Francisco and other Bay Area cities than between San Francisco and, say, New York .
We can also see that there is a great range of elasticity figures between studies, and therefore choosing an exact average of elasticity figures may be implausible. But Bartik posits that the "true" average effect of a 10% reduction in state and local taxes is likely to be an increase in economic activity of somewhere in between one and six percent, a long-run elasticity of between -0.1 and -0.6.
• Wasylenko (1997). It was once widely thought that local taxes had little impact on growth and firm-location decisions, but Wasylenko identifies a strong consensus to the contrary: Given recent empirical evidence, "it is increasingly difficult to argue that business climate [including taxes], however broadly defined, does not influence interregional firm locations."
Wasylenko also recognizes the greater importance of taxes at the intra-regional level, asserting that elasticities within a region appear to be at least four times greater than those at an inter-regional level. He also finds that reductions in business taxes tend to attract capital-intensive firms, which pay higher wages and benefit those with more job skills.
• Fisher (1997). Fisher's literature review, published for the same conference as Wasylenko's and as a complementary piece, focuses on the growth impacts of public services rather than tax rates. Fisher finds that the provision and maintenance of transportation infrastructure has the strongest link to growth by firm location . He also finds that public safety (police and fire) has a strong but less consistent effect on firm location, whereas education spending demonstrates the weakest effect.
• Mark, McGuire, and Papke (1998). "Authors of intra-regional studies more consistently find a significant effect of tax differentials [different tax rates between cities] on local economic activity [than those of inter-regional studies]. However, in both types of studies, when taxes are statistically significant, the size of the effect has not been large. Other factors, such as labor costs or labor quality, tended to be more important. These empirical findings have been supported by evidence gathered in surveys of businesses, which consistently place taxes low on the list of critical factors in location decisions..."
"...A different conclusion from this literature is that there is a role for publicly provided services in shaping a conducive environment for business. Many of the studies examined both taxes and government expenditures and found that spending on education, highways, and other types of services likely to be valued by firms has had a positive effect on economic activity. Indeed, some studies found that an increase in taxes coupled with an increase in spending on desirable services is a net plus for economic development."
Summary of Elasticity Studies
A look at the specifics of a number of studies is necessary to understand the detail and variety of the individual econometric studies.
While results vary, even within the subcategories of inter- and intra-regional studies, it is important to note that the studies were not uniformly "robust" with respect to changes to specification, time period, and measurement. The mixed results may be a result of the mix of industries studied-it may be that a spending variable is a positive factor for manufacturing, but a negative factor for wholesale trade. They may also be related to method-for instance, a tax variable may be a negative factor if the dependent variable is measured one way, but not a factor if it is measured another (Mark, McGuire, and Papke 1998) .
Also, there is a practical caveat to the empirical results on taxes and spending. While the regression results may indicate that a cut in taxes would result in a statistically significant but small boost in economic activity, the result depends on holding all other factors, including government spending, constant. For state and local governments facing balanced-budget rules, such an experiment is not feasible (Mark, McGuire, and Papke 1998) .
The following charts, based on those found in the report of the District of Columbia Tax Revision Commission (Mark, McGuire, and Papke 1998) , provide a quick look at the major studies that examine the impact of taxes on growth. For each study, the growth variable tested, the geographical unit of comparison, and the study's conclusion are indicated. For easier reading, a plus sign ( + ) has been added to designate studies that find taxes matter, a minus sign ( - ) indicates that taxes did not matter, and a combination of the two ( +/- ) indicates that results were mixed. Full citations for each study can be found in the bibliography of this report.
Inter-Regional Studies-Comparing one region (state, county, or metro area) to another.
+ Bartik (1985)
+ Mofidi and Stone (1990)
- Carlton (1983)
+ Dalenberg and Partridge (1995)
+ Papke (1991)
+ Helms (1985)
+/- Tannenwald (1996)
+ Hines (1996)
+ Wasylenko and McGuire (1985)
Intra-Regional Studies-Comparing areas to one another within a single region.
+ Charney (1983)
+ Luce (1994)
- Erickson and Wasylenko (1980)
+ McGuire (1985)
- Fox (1981)
+ Wasylenko (1980)
( + taxes matter, - taxes do not matter, +/- mixed results)
Other Important Findings
Studies with findings relevant to this report, but that did not fit in the above categories, are cataloged here.
• Luce (1994) finds, in a study based on the Philadelphia metropolitan region, that in terms of firm-location decisions, increased government spending on services does not fully compensate for high tax rates. He writes: "The effects of local tax increases on employment in specific sectors can be at least partially offset if the higher taxes lead to properly targeted additional spending. However, in this sample, the net effect on total employment will be negative except in very special circumstances." This means that cities counting on making up for high tax rates with high service levels should be cautious not to take false comfort-service improvements enacted with the purpose of attracting or retaining jobs should be systematically targeted to do so and must be perceived by businesses as valuable.
• Schmenner, Huber, and Cook (1987) find that corporations that say taxes are important to their locational decision-making appear to behave consistently with that stated preference.
• Studies by both Wasylenko and McGuire (1985) and Testa (1989) find that manufacturing location decisions appear to be more sensitive to taxes than non-manufacturing location decisions. We would expect this pattern because manufacturers tend to be capital-intensive, and most state and local business taxes are taxes on capital. For largely de-industrialized cities like San Francisco , this both reinforces the belief that manufacturing will not return and softens the negative impact of high taxes in terms of attracting and retaining non-manufacturers.
Local Revenue Hills
Many recent econometric studies have gone beyond simply finding the elasticity of growth with respect to tax rates. A core of literature has emerged that connects these elasticities with overall tax revenues to a jurisdiction. The result is the concept of the revenue hill , an idea that holds that "the positive effect on revenues of an increase in rates is significantly reduced by the negative effect on revenues that follows from the loss in the tax base" (Inman 1992) . The term "hill" refers to the shape of the revenue curve this creates-a curve depicting revenue increasing with tax rates until a certain point, peaking, and then falling off. This is the urban version of the infamous "Laffer curve."
What is the "Laffer Curve" ?
Economist Art Laffer, an advisor to President Reagan in the early 1980s, devised a curve for total tax revenue as a function of tax rates:
T* = revenue-maximizing tax rate
Laffer suggested that as taxes increase from fairly low levels, tax revenue received by the government also increase. However, as rates increase people do not regard it as "worth it" to continue working and eventually work less, quit, or move to other jurisdictions. This leads to a fall in income and therefore a fall in tax revenue. The logical end-point is with tax rates at 100% where no one would bother to work because all of their income would be taxed. As a result, tax revenue becomes zero.
It was often said that Laffer and his curve had been "discredited." To the contrary, the concept of the Laffer curve is now widely accepted. It was Laffer's assertion that national tax rates in the 1980s were to the right of the maximum point-and that it was the subsequent policy implications of this assertion, the 1981 federal tax cuts, that were discredited.
The existence of "revenue hills" means that taxation of a real economy is more complicated than a simple accounting relationship. As a matter of simplistic fiscal accounting, tax revenue flows from taxing some tax base such as income or property value at a chosen tax rate:
Tax Revenues = Tax Rate x Tax Base
The accounting relationship is linear, a straight line. In other words, if you multiply the same tax base by a higher rate, you get more money. But as a matter of fiscal economics, the size of the tax base is partly a function of tax rates. That is, the economic relationship between revenues and rates allows the tax base to change as tax rates are increased or decreased:
Tax Revenue = Tax Rates x Tax Base(Tax Rates)
The economic relationship is a convex curve. Conceptually, the two curves look like this:
Raising tax rates reduces the tax base because as rates increase, a higher proportion of the activity of individuals and firms are taxed, reducing the marginal benefit of "doing more." For example, the real world implications of raising business taxes are that some firms will choose to shut down, relocate, or scale back operations.
Beyond the mere existence of "revenue hill" curves, economists are also interested in knowing the shape of the curve, answering the question "how much is too much?" To do this, they estimate statistically the effects of tax rates on the tax base from the past responses by taxpayers to changes in rates. Based on empirical data, a part of the curve can be constructed; the remainder is then extrapolated.
The policy implications of revenue hills depend on "where" in relation to the hill's peak current tax rates lie. If they lie to the right of the peak, decreasing taxes should actually increase overall revenue; if they are located to the left of the peak, raising tax rates still increases revenue. Haughwout, Inman, Craig, and Luce (2000 and 2003) estimated the current rate position along revenue hills for Houston (property tax), Minneapolis (property tax), New York City (property, general sales, and income taxes), and Philadelphia (property, gross receipts, and wage taxes). They found that in every city except Minneapolis , tax rates were at or near the revenue hill peak. This means that Houston , New York City , and Philadelphia had reached their maximum potential revenues from the perspective taxes, at least with respect to variable tax rates. It is not known where San Francisco is along its "revenue hill."
The revenue hill curve implies that a tax increase is not an effective way to increase tax revenues when a city is at the vertex of the curve or to its right. At that point, described by economists as the point at which tax rates reach "unitary elasticity" with respect to revenues, the only way to grow tax revenues is to grow the economy. This might be accomplished through economic development policies, especially those that focus on enhancing firms' productivity, as these will not have the direct revenue impacts of tax incentives.
It should be noted that reaching the peak of a revenue hill-the maximum point for revenu- is not necessarily the optimal point for a city. Ultimately it is a question of the values of the citizenry whether the goal is to maximize the provision of public services like transit, parks, and education, or to minimize the tax burden.
Real-World Firm-Location Decisions
Generally speaking, studies that look qualitatively at how firms make individual location decisions find that while certainly not ignored, taxes play a relatively minor role in the decision and often only at the margin, once a variety of other decision screens have been applied. These studies include surveys of corporate executives and others responsible for making such decisions, as well as the comments of leaders in economic development.
Much work in this category gives cause for only very moderate recognition of tax factors in location decisions. William F. Fox, Professor of Economics at the University of Tennessee at Knoxville Business School, finds that generally tax incidence-the distribution of taxes among types of firms and among different taxable activities-can be as important to a firm's location decisions as tax rates (Fox 1997) . (We discuss this issue, of how business taxes are structured, directly in Part II of this report.) Economist Robert Lynch finds in his examination of actual firm decisions little evidence that the level of state and local taxation figures prominently in business-location decisions (Lynch 1996) .
While it is still true that personal factors such as the birthplace or preference of a company's CEO play a large role in a firm's location, locational decisions have in general become far more rational than they once were. This is particularly true for larger companies.
It is important to remember that location decisions are seldom only a location decision-they are usually part of a larger corporate planning process. Locational decisions can be sparked by a strategy of expansion, market repositioning, even contraction and downsizing. Because of this context, location decisions inherit a lot of criteria and dictates from larger-scale company plans. A location strategy is rarely a free-standing strategy created of its own accord.
It is also important to recognize that location decisions are conducted very differently depending on the size and organization of a company. Corporations with a centralized staff will generally form a team of representatives from various effected areas (e.g. transportation, distribution, logistics, personnel, real estate, and planning). On the other hand, companies with strong divisions and weaker corporate staffs normally carry out the location selection at a divisional level (Blair and Premus 1987) .
Blair and Premus outline the general design of the formal locational process within large, multi-plant companies. While the specific subject of their discussion was the location of manufacturing plants, their description can be considered more broadly as a foundation for understanding firm-location decisions in general.
1. Must-and-want list. The site selection team will develop a list of characteristics that are important for the location of a new facility based on factors economic and non-economic, quantifiable and non-quantifiable. The role of the plant in overall corporate strategy will also be considered. Desires to penetrate a new market, to segregate or integrate corporate functions, or to increase a firm's visibility may be elements of the decision. Usually, items on the must-and-want list are weighted to indicate which features are most important.
2. Gather and compare. The next step is to gather information about potential sites and to compare the features for each side against the must-and-want list. Sites will be eliminated in rounds as more detailed information is gathered.
Early rounds seek to select broad regions at a large scale; later rounds look at specific communities and locations. In other words, there is a tiered decision-making process-at first, regions are selected, and then individual cities . In selecting a broad region, the site selection team will focus on labor, state tax variables, climate, proximity to markets, and other features that have significant inter-regional variation but are similar within the region. Location factors applied for the smaller, city- and site-level analyses are those that vary within the region, including land costs, access to major roads, and schools.
What is a region?
The word "region" has many different meanings across academic disciplines. A region may be loosely defined as an area that shares some degree of social, political, economic, and environmental unity-This may be as small as the built-up Bay Area or as large as all of Northern California , the modern day version of the medieval Italian city's contado (the countryside within its sphere of influence).
But for the purpose of this report, we will consider a region to be a "commute shed," an area in which the same residents commute to the same range of jobs. This is the most relevant definition because the characteristics of a labor pool are essentially uniform for a firm located anywhere within a commute shed, because that firm is capable of drawing workers who live anywhere in that shed. This definition also corresponds roughly to the Census Bureau's definition of the Metropolitan Statistical Area (MSA).
3. Development processes begin. As specific sites are considered, the early, tentative stages of the building development process may begin. Discussions with local public officials regarding potential problems and incentives commence. If the company is looking to build a new structure in its new location, preliminary estimates of construction costs will be developed for inclusion in the corporation's capital budget.
It is hard to know how widely applied are strategic decision-making processes such as those described above. One way to determine this is to survey a large pool of corporate decision-makers as to their location-decision process and the factors that weigh most heavily in them. While such studies were common in the 1970s and 1980s, few of note have been conducted in recent years. Additionally, such studies do not elicit precise enough answers to settle the issue. We would like information such as whether the firm would have changed its decision if state and local taxes at the chosen location had been 10% higher, but such experiments are impossible to conduct (Bartik 1992) . Because of these problems, the older, survey-based studies will not be reviewed here, but the reader can reference the appendix for an interesting, if dated, literature review conducted by Blair and Premus (1987).
The Role of Consultants
Perhaps the most enlightening articles are those written by location consultants hired by companies to conduct location, relocation, and expansion studies. Consulting firms can usually conduct a site analysis more cheaply than in-house staff because they have access to data and experience in the field. As of the late 1980s, consulting firms were used in about one-third of the locational decisions of Fortune 500 companies (Blair and Premus 1987) ; today the figure is probably higher.
It is useful to look at the work of consultants because they have devised structured, formulaic approaches to location decisions that, in varying forms, have been widely applied. According to Blair and Premus (1987), consulting firms first assist clients in understanding the purpose of the new plant or office and in determining any important locational factors, the "must-want list." Based on this list, a systematic process identifies a small number of regions that meet the general criteria. Consulting firms have large data banks and computer models that assist in this process.
But the location decision is not all science; Blair quotes Wardrep (1985), who writes that during the "initial stages of the location decision[,] the nonstandard nature of the location search process, and the complexity of the elements that need to be considered, combine to create hesitancy in many managers."
A more detailed explanation of a location consultant's strategic process is offered by Robert M. Ady, Executive Consultant for the Fantus Corporate Real Estate Group of Deloitte & Touche. Ady consults on firm relocation and regional economic development issues and has written on the strategy he applies in advising companies on location, relocation, and expansion decisions (Ady 1997) . Simply described, the stages of his decision-making process are as follows:
1. Broad area of search. In the earliest stage of a location decision, a broad area of search is defined. Usually, "broad area" means a list of several metropolitan regions. Tax considerations are very minor at this point; the focus is instead on wage and transportation differentials as well the relative quality of the workforce.
There is also consideration of potential "fatal flaws"- specific attributes of a place that make it particularly unsuited to a particular firm. For instance, an aircraft manufacturer would be discouraged from locating in regions with hard winters, because it would want to maximize good flying weather. High tax rates would prompt elimination at this stage probably only if they are considered a unique "fatal flaw" for a certain company.
2. Cost comparisons for specific locations. For this step, the field of potential locations has been narrowed down to a number of specific jurisdictions in one or two metropolitan areas. Estimates of operating costs are created for each location and compared. Because taxes comprise only 4-5% of "geographically variable operating costs," a much higher weight is placed on labor, transportation, occupancy (space), and even utility costs. In general, when considering locations across broad geographical areas, other cost variations are almost certain to "swamp" the effects of tax differentials on firm location.
Public services-the expenditure side of taxation-may nevertheless be significant. For instance, spending on highways and transportation can affect transportation costs associated with specific sites and expenditures on education can affect labor costs. But the importance of public services varies considerably across sectors, industries, and firms, and may indeed depend on the preferences of a firm's top management.
3. Detailed calculations of costs. Only a handful of locations in a limited geographical area-usually limited to a particular part of a single metro area-proceed to this step. Detailed accounting of all costs is conducted, including yield "net" taxes that take into account all taxes, abatements, and incentives. It is at this stage that tax differentials can play a significant role and larger companies begin talks with cities for incentive packages and custom-designed tax breaks.
Despite formulas such as this, the complexity and informality of real-world firm-location decisions cannot be forgotten. Some companies conduct detailed accounting analyses of each location under consideration, while others simply follow intuition or the desires of a CEO. For instance, some companies choose not to consider moving away from their "home city" because staying offers intangible civic or personal benefits.
But this does not mean that many firms' location decisions are grossly irrational. To the contrary, "the majority of businessmen... are familiar with the operations of their industry, know where their markets are and where their raw materials comes from, and from this general knowledge can pick the likeliest spots and investigate them further" (Alonso 1975) . Alonso adds that "the 'survival of the fittest' will mean that, however the decision is made, it will be those industries that are well located that will survive and become important."
On the other hand, political uncertainty frustrates the use of taxes as a rational part of a location decision. Pomp (1985) argues that because most relocating companies choose to stay at their new location longer than any group of elected officials is likely to stay in power, "current tax levels, special concessions, or special features of the tax law may not be a reliable basis upon which to make a multi-million dollar investment." And of course, political uncertainty has a negative effect beyond tax issues: businesses dislike politically uncertain situations in any context because it makes it difficult to forecast costs.
In sum, it is apparent that when robust strategies are applied, taxes matter primarily in the final stages when only a few locations that share most other characteristics are under consideration. Salary and labor-force quality between cities in the same region are very similar, so only rent and taxes represent a marginal cost. In most cases, this means that taxes matter most within a region , when a company has already decided to locate in a certain metropolitan area. This is when taxes may help decide which city within that region gets the jobs.
Considerations for Central Cities and Downtowns
Although taxes may matter and San Francisco does have high business taxes compared to other Bay Area jurisdictions (see the tax data in Section III), we cannot automatically expect high tax rates to cause the attrition of firms from the city. While many of the non-tax factors in location decisions are essentially constant among nearby suburban jurisdictions (most significantly access to the same labor pool), the central city enjoys a competitive advantage, at least for certain types of firms. This advantage dampens, to an extent, the harsh realities of intraregional tax competition for San Francisco . As a top-tier central city with a famous downtown, San Francisco can, in a sense, "get away" with charging higher taxes, at least in the short run.
A few high-profile studies have shed light on the city's advantages in general. For example, Antonio Ciccone and Robert Hall have established a direct correlation between urban density and productivity (Ciccone and Hall 1996) . The result of this productivity advantage is that the competitiveness of central cities is enhanced, compensating to some extent for their disadvantages, one of which is higher taxes.
In fact, several studies have found that tax differentials matter less for central cities . Studies by Mills and Price (1984), Bradbury, Downs, and Small (1982), and Grubb (1982) do not find any significant effects of a central city's relative tax rates on its share of a metropolitan area's employment. This is plausible because central city locations in general may not be good substitutes for suburban locations (Bartik 1992) . Indeed, most of the data used in the econometric studies summarized above were dominated by observations on individual suburban jurisdictions; the large tax effects they indicate should be interpreted as applying primarily to suburban jurisdictions. Cities are special.
The Advantages of Downtown
A central city location-and particularly a location in a Central Business District (CBD)-offers certain tangible and intangible benefits to companies that boost productivity, reduce costs, and enhance the available labor pool. These benefits, centralizing forces that contradict the supposed decentralizing effects of modern transportation and communications technology, help mitigate the magnitude of the central city's tax burden:
• Face-to-face interaction. While it was projected that telecommunications technology would make physical location irrelevant, it has by now been widely reported that the need for face-to-face interactions in high-level business functions has not diminished. Teleconferencing and Internet communications are still far from capable of reproducing the unique sensation of "being there." "Telephone calls, faxes, and e-mail messages are, by their nature, planned, reaching only the people they're intended to reach. Cities are unparalleled in their ability to provide the constant, spontaneous interchange of knowledge and insight in which new ideas thrive" (Glaeser 1996) . In fact, technology may have actually heightened the need for face-to-face encounters as the overall volume of communication has increased drastically. The dense concentration of firms in central city locations such as downtown San Francisco facilitates such face-to-face interactions, from formal meetings and lunches to unscheduled encounters on the street.
• Employee and executive lifestyle. The highly educated workforce (and especially management) of today's leading firms, the much-ballyhooed "Creative Class," demands social and cultural amenities that are, as a rule, not widely available in suburban locations (Florida 2002) . These amenities include proximity to theaters, museums, art galleries, restaurants, and shopping, as well as a general sense of being in an important place, a great city. Simultaneously, the demographics of the American family-particularly the well-off family--tend toward smaller households with fewer children and more disposable income. As a result, the kid-friendly, suburban American Dream has been joined by the dream of an urban lifestyle in which cosmopolitan cultural experiences are highly valued. On a less profound level, simply meeting friends for dinner after work is easier in a dense downtown than in a more dispersed environment. In fulfilling these lifestyle demands of the educated middle-class, cities like San Francisco excel.
What is the creative class?
Richard Florida, Professor of Regional Economic Development at Carnegie Mellon University, argues in his book, The Rise of the Creative Class (2002), that a new social strata of highly educated, innovative workers-scientists, engineers, architects, educators, writers, artists, and entertainers-have become critically important to the growth of high-performing regions in the developed world. Florida finds characteristics such as creativity, individuality, diversity, and left-of-center politics common among members of this "class." Creative people seek, along with the firms that employ them, places that are culturally active and tolerant of diversity, and therefore find older, denser central cities attractive.
The San Francisco Bay Area is ranked 5 th of regions with over one million as a center for the "Creative Class."
• Supportive business services. The dense concentration of firms in downtown locations is generally well suited for small, specialized business services firms and the companies that purchase their services. By now the tendency of large companies to contract out activities that were previously conducted "in-house" is well known. The services of attorneys, designers, consultants, engineers, market analysts, public relations staff, and human resources managers are now likely to be purchased from outside, specialized firms. In an environment where business services are decentralized and ubiquitous, it is impossible for the buyers of these services to have perfect information, to know if they are getting the best deal. As a result, buyers of business services must rely on trust -which is most easily built with face-to-face interactions. The concentration of firms that provide services to other businesses, from lawyers to graphic designers to computer programmers, makes it easier for businesses in places like downtown San Francisco to network multiple firms into a "virtual corporation."
• Accessibility. Although the old idea of the metropolis as a central city from which spokes of suburban development radiate has been overlaid with a "network" pattern of many links and nodes, the central city still enjoys a privileged position in terms of accessibility. Because of geographic position and historical development patterns, downtowns are often at the center of highway systems and public transport networks. Compared to suburban locations far to one side of the region's center, central cities have a wide "catchment area" for workers and potential customers, usually the entire population of a region. Because the most accessible, most central point in a network is also often the most congested, this advantage is strongest, of course, where transportation "lines" (highways and transit) have enough capacity to efficiently move people.
• Information sharing. The exchange of information is facilitated in downtown environments. For example, the need to quickly dispatch sensitive legal documents still remains important and is facilitated by downtown bicycle courier services that can deliver packages in minutes. While there has been speculation that all legal documents will eventually be handled electronically and signed "virtually," so far there is still the need for hard copies of contracts, agreements, and lawsuits. Additionally, when digital networks come under fire for security breaches, hard delivery becomes once again attractive to privacy-minded managers. Densely packed downtowns with courier infrastructure, like San Francisco 's, readily accommodate these needs.
• Corporate prestige. Downtowns, and especially the downtowns of top-tier, internationally recognized cities, provide a level of prestige and public recognition that suburban locations do not. This is justifiably important for publicly held companies because a visible location may "advertise" this prestige to potential investors, both the general public (which goes downtown as tourists or to eat and shop) and brokers in nearby investment firms. In cases where companies have built distinctive structures downtown, the location advertises to an even broader audience. For example, the profile of San Francisco 's Transamerica Building , as part of the Transamerica Corporation's corporate logo, can be spotted worldwide on shipping containers and has become virtually synonymous with the city's skyline. Even for private companies, the cachet of certain urban addresses is alluring, and San Francisco 's Financial District remains one of the premier locations in the United States .
• Close to government. Many types of firms must be close to governments and government agencies, and for historical reasons, downtowns are often home to concentrations of such public-service functions. For firms that interact extensively with government, proximity allows quick travel, face-to-face interaction, and heightened lobbying visibility. Law firms often choose to locate near courts; public utility companies often have their headquarters near to the offices of the responsible regulator. San Francisco is an important center of government at several levels and has a high concentration of state and federal courts, including the chief location of the Ninth Circuit Court of Appeals.
In sum, urban downtowns offer special advantages that compensate, to an extent, for their higher costs. As a city for business, San Francisco faces an extreme cost hurdle, but some firms will overcome that. But it is not enough to simply let these advantages console the high-tax city that not all firms will want to leave; the effects of these advantages must be carefully nourished and promoted, as well as marketed to resident and prospective firms.
Finally, it should be noted that the facilitation of the downtown advantages described above is often a political choice undertaken for the benefit of high-level business functions at the expense of other types of economic activity. This is especially true for San Francisco : in the 1960s and 1970s, city policy consciously guided a major shift in the city's economic base (from manufacturing and distribution to high order services), capital stock (from low-rise to high-rise), and labor force (from blue collar to white collar). Indeed, land-use and other public-policy issues all play an interconnected role in determining a city's economic makeup.
In a discussion of business taxes, then, it is important to recognize that as a component of business costs, taxes can also have a tremendous impact on what type of economic activity flourishes in a city . Just as other policy and planning decisions have favored high-level, white-collar firms, increasing business taxes in San Francisco may squeeze out economic activities that cannot pay these costs, creating further specialization in the San Francisco economy.
A Footnote: The Political Purposes of Local Tax Studies
In considering the policy implications of the studies reviewed here, it is necessary to recognize the political bias of the various researchers, to recognize what the policy implications of a study were meant to be. In addition to the more general methodological categories which organized the previous section, a division of "purpose" can be discerned: Some studies want to show that tax cuts are a costly and only marginally effective means of creating growth, whereas others aim to "warn" high-tax jurisdictions that their policies may be driving business from their boundaries.
These distinctions of purpose complicate the interpretation of the literature in the context of San Francisco 's situation. If lowering taxes is a costly and ineffective way to increase economic activity, it obviously does not follow that increasing taxes is the best policy choice. Likewise, warnings that high-tax cities should "watch i" are of little help, especially because cities must be concerned with every variable that goes into the budget. That is, they must also be concerned with levels of public services and community concerns.
Implications of the Literature
By this point, we have looked at a wide variety of studies. Most suggest a negative impact of taxation rates on firm-location decisions, but many qualify this impact. What can we take away from this discussion?
First, tax differentials matter much more on the intra-regional level than the inter-regional level. For instance, taxes matter more when a firm is deciding between locations within the Bay Area than when deciding whether to locate in the Bay Area or the Seattle area. This was confirmed by the elasticity studies discussed above.
Second, taxes do matter, and there is a theoretical point at which they discourage economic activity such that total tax revenues actually fall. However, it is very difficult to pinpoint at what tax rate this point is.
Third and finally, dozens of non-tax factors also influence firm-location decisions and the health of a city's economy. We must not be myopic and focus too heavily on taxes.
In frustratingly-but necessarily vague-terms, the bottom line is that taxes play a role in economic development, but it is not clear how big. Policymakers must appreciate the complexity of taxes' influence on local economies in order to make informed decisions.
II. Which Taxes?
If the previous section asked how high overall taxation levels should be, this section will examine the question of which taxes are best . In other words, it will examine the policy choices San Francisco faces about who to tax and what activities to tax.
Under the question of who to tax, the choices include the relative burdens that fall on: individuals vs. businesses; small businesses vs. large businesses; and firms in one industry type vs. firms in another industry type. Currently in San Francisco , only the largest 15% of businesses pay considerable business tax to the city.
Under the question of what activities to tax, sometimes referred to as "tax incidence," the choices include: taxes on production (e.g. payroll taxes and gross receipts) vs. taxes on consumption (e.g. water and energy) vs. taxes on property. Taxes on production, by definition, fall only on businesses, whereas taxes on consumption and property fall on both individuals and businesses.
When considering which activities are to be taxed, it must be remembered that taxes discourage the activity on which they impose costs while potentially encouraging substitute activities. A tax on energy consumption will encourage energy conservation. A tax on gasoline will discourage driving or encourage more fuel-efficient cars. A tax on hotel occupancy will discourage hotel reservations. And a tax on payroll will discourage employment.
Conversely, tax breaks on certain activities will encourage them. The mortgage tax deduction encourages people to invest in home ownership and leads people, on average, to "consume" larger amounts of housing relative to other things they can purchase on credit, which are not tax exempt. Tax-deferred retirement plans encourage people to save money.
As with the previous discussion on how high taxes should be, there is no clear single answer for which taxes are best. But it is clear that some types of taxation have advantages over others, and these relative advantages are important to the debate over business taxes.
Two Principles of Taxation
The Ability-to-Pay Principle. The ability-to-pay principle suggests that the total tax burden be distributed among individuals and firms according to their capacity to bear it, taking into account all of the relevant personal characteristics in such a way that the relative loss in economic capacity resulting from the tax is equal between all individuals or firms.
The Benefit Principle. According the benefit principle, taxes should work the way prices do in private transactions; that is, they determine what services the government provides and asks the user to pay for them. This idea is known colloquially as "pay for play." But assuming that income does not correlate with need for services, the benefit principle may be regressive, conflicting with the ability-to-pay principle.
The Role of Business Taxes
The first choice governments face is whether to collect taxes from individuals or businesses. Some cities choose not to have business taxes at all. For instance, Palo Alto does not have business taxes, although it has higher utility taxes than most jurisdictions. However, businesses are generally identified as the best organizing unit through which individuals can derive the benefits of economic activity as consumers, workers, and shareholders; this means the firm is a particularly logical and "central" level at which to tax.
Additionally, because business taxes capture economic activity generally rather than the derivative activities of residents (personal income, spending, etc.), business taxes can capture revenue that taxes on individuals cannot, making them uniquely effective. In an open economy where people, businesses, goods, and services cross jurisdictions, a portion of the sales and income that originate within one location represents sales to nonresident customers and income paid to nonresident owners and shareholders who cannot be taxed directly. Consequently, taxing businesses may not only be administratively convenient, but the only way a local government may derive some tax revenues from those individuals who are not residents of the jurisdiction, but still derive some economic benefit or make use of its publicly provided services and infrastructure. The high number of tourists and non-resident commuters who come into San Francisco make business taxes particularly important for this city.
But even though business taxes are collected from businesses, their economic burden is ultimately borne by people :
• Owners and shareholders, if business taxes lower equity returns
• Employees, if business taxes result in lower wages or reduction in employment
• Customers, if business taxes raise prices
• Property owners, if business taxes lead to lower property values
The citizenry can decide that for any given tax, the people who will be affected can afford it, or that the benefits outweigh the costs. But it is important to keep track of the human actors behind the curtain of "a business."
Constraints on Local Taxes in California
While it is certainly useful to discuss the merits of different types of taxes, the ability to change or raise taxes is considerably restricted in California . Tax policy choices available to a municipal government are severely limited by the voter anti-tax movement, which has won significant victories with passage of Propositions 13 in 1978, 62 in 1986, and 218 in 1996:
Proposition 13 froze property tax rates and required voter approval of special-use taxes, which were formerly levied by local legislatures. The property tax rate is limited to 1% of taxable value plus a rate necessary to pay of indebtedness such as bonds authorized by the voters. Property can only be assessed when a change in ownership or major construction occurs.
Proposition 62 greatly expanded the mandatory voter approval of tax increases in most jurisdictions, requiring a majority vote on general taxes.
Proposition 218 added even more limitations on taxation and required voter approval of property-related assessments, fees, and charges. In addition, Proposition 218 extended the limitations on revenue collection to all California jurisdictions, including charter cities such as Los Angeles and San Francisco , which some argued were outside the reach of Proposition 62.
Together, these three initiatives placed unprecedented limits on the ability of local governments to levy taxes and generate revenue. This makes the tax structure inflexible, as all changes must go to the voters. There is a resultant hesitancy to try new things, as it is difficult to make minor adjustments to a tax structure if it is not right on the first try.
Source: "Crisis and Opportunity in the City Budget," SPUR's February 2003 newsletter.
Types of Business Taxes
Businesses pay many types of taxes, including but not limited to those commonly referred to as "business taxes." Businesses also pay a variety of other taxes covering licensing, permitting, utility consumption, and property taxes. While telephone and parking taxes may be significant in some jurisdictions, this section will focus on those that make up the largest chunk of tax revenues that come from businesses: payroll taxes, gross receipts taxes, property taxes, and valued-added taxes. Sales taxes will not be discussed, as their costs are almost always passed down to the consumer.
The Payroll Tax
"The Payroll Expense tax is a tax on the payroll expense of persons and associations doing business in San Francisco . All businesses that engage, hire, employ, or contract with individuals, as employees, to perform work or render services within San Francisco are subject to the Payroll Expense tax. Payroll expense is the total compensation paid, including salaries, wages, commissions, and other compensations to all individuals who, during any tax year, perform work or render services in whole or in part in San Francisco." (Crapo 2001) . San Francisco currently levies a payroll tax of 1.5%.
Payroll taxes are fairly uncommon. According to Michael Coleman, an economist for the League of California Cities, only 15% of municipalities in the state rely on a payroll tax to provide a form of business-tax revenue. And as the tax-comparison data included in Section III of this report show, payroll taxes are relatively uncommon across the country. Where they are found, payroll taxes are usually associated with older central cities. The rarity of payroll taxes can be seen as a disadvantage, as they may be unfamiliar to some firms and they can create problems for firms attempting to compare tax burdens between cities.
A payroll tax may also have particularly strong negative effects on employment . The negative effects of payroll taxes on employment include:
• Firms seeking to reduce their payrolls through wage cuts, layoffs, and net attrition and relocation of the workforce.
• Discouraged growth, as there is an effective disincentive for firms to expand their payrolls. Companies are likely to move their growth divisions out of the city, and outside firms are less likely to open a branch office in a city charging the payroll tax.
• Incentivized use of labor-saving technology, which can permanently reduce employment.
• Further skewing the available jobs toward white-collar work requiring highly educated workers. This is because high-level service firms are generally not labor-intensive, and so the payroll tax weighs less heavily on them.
The negative employment effect of the payroll tax was witnessed empirically in the case of Philadelphia, where 165,000 jobs were lost between 1970 and 1977, widely thought to be due in large measure to an increase in payroll taxes of 43% (Crapo 2001) .
There are several features about the payroll tax that have particular effects on San Francisco :
• Biotech. Although the Bay Area hosts the largest concentration of biotech firms in the world and University of California at San Francisco (UCSF) has spun off many biotech firms, few biotech firms have located within San Francisco . However, dozens of biotech firms have located immediately adjacent to the city in South San Francisco . The paucity of biotech firms in the city is partly a result of how the payroll tax treats biotech's unique business model. Typically, biotech firms operate for ten to fifteen years off of invested venture capital before their drug research pays off in the form of a marketable product. With no revenues, these firms are very cash-poor and cost-sensitive. It is, therefore, extremely onerous for them to pay payroll taxes on the salaries of employees at a time when the firm has no income. At the same time, biotech firms pay very little tax in almost every city in the Bay Area except San Francisco . One economist reported that a biotech firm choosing to locate in San Francisco can expect to pay between 18 and 800 times as much in business taxes than it would have had it located in another Bay Area community.
• Partnerships. One of the quirks of San Francisco 's payroll tax is in a business partnership, the income of a firm's partners is not taxed as payroll because it is considered earnings from an ownership interest in that company. This means that law firms, dental and medical offices, architects, engineers, accounting firms, and consulting firms organized as partnerships are taxed less heavily than firms not organized as partnerships. This translates into $10 million to $15.8 million of lost revenue for the City. (Of course, partnerships are still subject to federal and state income taxes.)
Despite its many drawbacks, however, there are certain benefits to payroll taxes. One advantage cited by economists is that that payroll taxes are easy to administer, as the tax is easily calculated for reporting purposes. Also, the payroll tax efficiently captures some of the economic activity generated by in-commuters and other nonresidents of the city.
The Gross Receipts Tax
The term "gross receipts" refers to the total amount charged or received for all sales of goods and services. As noted in the introduction to this report, San Francisco 's gross receipts tax was repealed by the Board of Supervisors on May 25, 2001 , in response to a lawsuit by major firms.
Gross receipts taxation is so common that in many places it is referred to simply as "the business tax." Gross receipts tax is the prevailing standard for urban business taxation in the US . It is probably true that gross receipts taxes have less of a direct, negative impact on payrolls through wages and employment levels than do payroll taxes. Instead, they weigh heavily on firms with high revenues per employee. The collection of gross receipts tax is more complicated than for payroll tax, especially considering that many cities levy very different gross receipts tax rates on different types of firms. And it should be noted that in Los Angeles , which now charges only a gross receipts tax, this tax is reported to be widely disliked by business executives.
Property tax is levied as a proportion of value of real property. Although property taxes are of course not exclusive to businesses, it can be a major cost for some types of firms, especially manufacturers. The San Francisco property tax rate for tax year 2002 (i.e. fiscal year 2002-03) was 1.117%, applied to the net assessed property value as determined by the County Assessor . Generally, the assessed value is the cash or market value at the time of purchase. This value increases not more than 2% per year until the property is sold or any new construction is completed, at which time it must be reassessed. After the Office of the Assessor/Recorder has determined the property value, the Office of the Controller applies the appropriate tax rates, which include the general tax levy, locally voted special taxes, and any city or district direct assessments. The general tax levy is determined in accordance with State law and is limited to $1 per $100 of assessed value. After applying the tax rates, the Office of the Controller calculates the total tax amount. Finally, the Office of the Treasurer & Tax Collector prepares property tax bills based on the Office of the Controller's calculations, distributes the bills, and then collects the taxes (www.sfgov.org/site/treasurer_page.asp?id=8098#1).
Property taxation is complicated, especially in California . In this state, severe controls on property tax rates and reassessment depress property tax revenues and distort the tax burden. Other more general factors make collecting property taxes peculiar: infrequency of reassessments and inequalities of assessment create uncertainties and inequity. Additionally, property taxes impact firms unevenly across sectors: property-intensive businesses such as manufacturing, wholesale warehousing, and real estate are heavily impacted, whereas business-service firms, which pay a portion of their landlords' taxes through their rent, are less-severely impacted.
Because of these problems and inequalities, property taxes are not usually considered an alternative to other types of business taxes, but rather a complement to them. They are also seen as part of a larger system of real property taxation that also affects non-commercial landholders and homeowners.
The value-added tax (VAT) is a broad-based tax levied on that portion--the "value added" portion--of the final product of a business that is over and above the value of the materials purchased. Each business is taxed on the addition to value it contributes to the final product or service. There are two methods of arriving at this tax base for a value-added tax: the deduction method and the addition method. Under the deduction method, the value added by any individual firm is equivalent to its total sales receipts less its costs for materials. The addition method bases the tax on the total of the firm's federally taxable profits with the addition of items that reflect the value added by the business that are excluded from federal taxation. These include the cost of labor, depreciation, and interest (www.crcmich.org/TaxOutline/glossary.html). San Francisco does not charge a value-added business tax.
Economists argue that value-added taxation does not distort behavior, but simply has a blanket effect on all economic activity. Additionally, because value-added taxation applies the tax only against the value "adde" by a firm, multiple taxation of the same business activity is avoided and transactions between businesses are treated the same as those between vertically integrated operations within a single firm. Added-value business taxes at the city level are uncommon, and are generally uncommon in the United States , in spite of their popularity with economists. This method of taxation is more commonly found in European countries.
Registration fees are levied on businesses in exchange for licenses or Business Registration Certificates giving them the right to operate. These fees are paid by all businesses, although they represent a greater burden for smaller companies, for which they are often the primary form of taxation. In San Francisco , business registration fees are based on a firm's payroll expense: Firms with a payroll expense of less than $66.67 pay a fee of $25, those with payroll of $66.67 to $666,666.66 pay $150, and those with payroll of $666,666.67 to $3,333,333.33 pay $250. For companies with a payroll greater than $3,333,333.34, $500 is charged (City Comptroller's Office).
A variety of more-specific taxes affect businesses. For example, in San Francisco , utility, parking, and hotel taxes affect businesses. Other cities have different combinations of such taxes. While detailed discussion of these taxes is outside the scope of this report, it should not be forgotten that they are part of the City tax revenues paid by businesses.
Which Type of Business Tax is Best?
This chart summarizes the principal impacts of the primary types of businesses taxes, including the advantages and disadvantages of each. The primary sources for this information are presentations by Margois (2003) and Rydstrom (2002), as well as Crapo's payroll tax report (2001) and interview (2003). More detailed discussions of the chart's contents follow.
Discourages high payrolls, discourages payroll and employment growth
Discourages high revenues (with low profit margins)
Discourages capital investment
No distorting effect, general dampening effect on economic activity
· Simplicity of administration on both ends
· Captures revenue from in-commuters
· The prevailing standard
· Integrated with larger system of property taxation
· Does not directly discourage economic activity
· No distorting effect on economic activity
· Avoids double taxation of the same activity
· Decreases incentive for existing firms to increase employment
· Encourages growing firms to move growth divisions out of the city
· Discourages start-up firms
· Discourages companies from opening branch offices in city
· Affects different types of firms disproportionately
· Often difficult to administer
· Reassessments of property value are infrequent
· Assessment inequalities can create uncertainty and inequity
· Very difficult to collect
· Uncommon in the United States
Labor-intensive firms such as restaurants and firms with local back-office functions
Revenue-intensive companies such as commercial real estate, auto rentals, gas stations, taxi companies, and contractors
Property-intensive firms such as manufacturers, distributors, and commercial real estate
If well-administered, effects all sectors equally
Capital-intensive firms, usually white collar
High-margin, low-revenue firms
Firms with no real property and low space needs
Taxes as a Way to Correct Pricing Externalities
It's important to note that there is a movement in some circles to charge taxes on activities that cause negative effects, such as the consumption of energy and water, rather than on activities that are beneficial to society, like creating jobs. Redefining Progress, an Oakland-based non-profit think tank, has been a strong advocate of this position. They argue that revenue raised from environment- and energy-based taxes could be used to reduce other taxes such as personal income taxes and business taxes (Hammond, Merriman, and Wolff 1999). This idea, which they call "tax waste, not work," has been endorsed by New York Times columnist and Princeton economist Paul Krugman.
The Concept of "Externalities"
Externalities are defined by economists as any market transaction in which the costs or benefits are not fully incorporated into the price of the transaction. There are both positive and negative externalities.
"A negative externality is defined as a cost imposed as a result of an activity on people who do not participate in the activity and who are not necessarily considered by the people participating in the activity." In the classic example, if a manufacturing firm is allowed to pollute the environment without paying for the costs of cleanup or pollution prevention, then the market cost of producing its goods will be underpriced compared to the true social and environmental cost of producing the goods.
A positive externality, by contrast, occurs when a market transaction has benefits for other people, which are not included in the price of the transaction.
Source: William Thomas Bogart, The Economies of Cities and Suburbs. ( New Jersey : Prentice Hall), 1998, p. 22.
Taxes can be used to help correct pricing externalities. For example, the consumption of gasoline has been recognized to impose many costs on society, the most obvious of which is air pollution. Taxing gasoline use at higher levels would help correct these negative externalities. That is, the user of gasoline will more fully realize the costs of his activities, and will likely reduce his consumption of fuel. But it must be remembered that taxes have cascading effects: that is, a gas tax will be directly borne by automobile and truck users, but will also be felt indirectly--in the form of higher prices--by anyone who buys products shipped by road.
Taxing the consumption of natural resources such as energy, water, and petroleum, rather than the far more positive activities of making money (revenue) and employing people (payroll), might discourage harmful activities while encouraging economic growth. As San Francisco policymakers consider their options, perhaps they can give some thought to the benefits of shifting the tax base "from work to waste," sending a message that is both pro-business and pro-environment.
The Role of Econometric Modeling
Some policymakers and scholars argue that the right way to choose the "best" taxes is through sophisticated, quantitative econometric modeling. Using complex models to look at the tax base and tax revenue, econometric modeling can be used to determine a favorable and equitable tax burden among sectors and forecast revenue and growth results.
For instance, Los Angeles , after spending five years on a major revision of its business tax policies, moved to a gross receipts tax with broad support from both business and labor. Through the process, L.A. streamlined its business tax code, simplifying its application across sectors by redefining the different categories that are charged different rates ("indexing") and reducing them in number. Many of these decisions were based on econometric modeling.
There are some serious problems with using econometric modeling, however. Because of volatility and uncertainty in the national and global economies, the predictions of econometric studies may not be received with much confidence. And as long as decisions on tax policies are ultimately made by politicians and voters; serious quantitative modeling may have little real say. But it is important to recognize that analysis of tax decisions shou ld go beyond simple arithmetic.
Beyond Economics: The Perception of Taxes
Regardless of what the economic models predict, there is another dimension to tax policy that San Francisco must consider--that the perception of business taxes may have an impact rivaling that of the taxes themselves.
The "Business Climate"
A general notion of how friendly a city is to business-widely referred to as"business climate"--may impact the firm-location decisions of businesses. The concept of business climate includes such issues as: What are local attitudes toward for-profit enterprises? What are the attitudes toward development? Is the city welcoming? It may not always matter how accurate this image of business climate is. Reputation can weigh heavier than fact.
Many people consider high business taxes an attribute of a city that is inhospitable to business. It is also possible that the reputation of a city as high cost may influence firm-location decisions even when the actual quantitative impact of the taxes is not prohibitive.
However, there are other important contributing factors to a city's business climate:
• The political activities of its population and elected officials, especially as measured in terms of highly visible events such as demonstrations. Will residents make firms look bad by protesting outside their offices?
• How quickly and predictably the city handles proposals for real estate development. Will it be able to expand when it needs to?
Unfortunately, San Francisco may be viewed as inhospitable to business because of these two reasons. High taxes, a notoriously left-leaning political scene, and the "development wars" may have set a tone of hostility toward business. But while it is neither possible nor desirable to most to move San Francisco 's politics toward the right, it may be possible to change the business climate.
In theory, San Francisco can have a good business climate despite high taxes and progressive politics: economic development initiatives that focus on boosting business productivity, in addition to marketing the productivity and labor advantages the city already possesses, may be able to recast San Francisco 's business climate as favorable. As many European countries have shown, high-tax environments with left-leaning populations can still be attractive places for business. But this would require a very different kind of leadership in San Francisco , one that could build on the underlying common ground between progressives and the business community. In order to get the high levels of public service that progressives want, there must be a healthy economic base to generate high levels of tax revenue.
III. Tax Comparisons
In light of the studies described above, it is important to look at how San Francisco 's business tax rates compare to those levied by other jurisdictions, in order to answer the question: how competitive is San Francisco 's business environment? This section of the report will look at data comparing business tax rates across the Bay Area, California , and select national cities.
What is a Comparable City?
Comparing San Francisco 's tax rates with those of other cities is not as simple as it may seem. Cities vary greatly in terms of their relationship to their greater regions and the structure of their economies. The notion of what is a "comparable" city depends both on the purpose of the comparison as well as governmental characteristics such as the geographical size of the city and the presence of overlying jurisdictions.
Two major notions of comparable cities could be:
• Other jurisdictions in the Bay Area. Reflecting the fact that business tax differentials matter more at the regional level, some would argue that comparing San Francisco to smaller, suburban jurisdictions in the Bay Area would be most relevant. These cities share with San Francisco many important attributes such as labor-force quality and housing costs, although tax rates vary wildly. Because they compete with San Francisco in actual firm-location decisions for investment and jobs, these cities are critical to the comparison--they allow us to see, in effect, how our "prices" compare with those of our competitors.
• Other central cities with downtowns. While San Francisco does not compete directly with other major central city jurisdictions such as New York City and Chicago after the first phase of firm-location decisions, these cities are more relevant comparisons, some might argue. This is because they allow us to benchmark San Francisco 's tax rates against cities that are more similar in terms of factors that drive costs, such as levels of unionization and the advantages of downtown, as well as factors that drive public spending such as the need for social services. Central cities almost always have significantly higher taxes than surrounding suburban jurisdictions, but some central cities have higher rates than others. Keeping in mind that as a central city San Francisco is manifestly different from smaller suburban communities, comparing the city with other major cities allows us to compare our tax practices with those of cities that share important characteristics.
Whichever type of city is seen as a more comparable case, many technicalities of government structure and the way taxes are collected further complicate comparisons. These complicating factors include:
• Extent of the city. This distinction is a result of historical accident: as some cities grew beyond their original boundaries, they annexed suburban areas; other did not grow their boundaries and their suburbs were incorporated separately from the central city. Phoenix is a city that annexed peripheral areas, whereas St. Louis and San Francisco are prime examples of cities that did not. Whether suburban areas were annexed to central cities helps determine the social and economic profile of the city-those that include more "suburban" areas within the same municipal government as the central city tend to have higher median incomes--as well as the size of the overall tax base. Small central cities surrounded by separately incorporated suburbs must face the costs of aging infrastructure and a needy population, while spreading these high costs over a relatively smaller tax base.
• City/County relationship. Similarly a product of historical accident, the relationship of a city to its county is important. For some cities, the city and county limits are identical, as with the City and County of San Francisco and the City and County of Honolulu . Los Angeles County , on the other hand, includes territory that is not a part of the City of Los Angeles . This difference matters because it makes it hard to compare combined city-counties with separated ones: where cities and counties have different boundaries, all the types and purposes of taxes that are levied together in city-counties are split between the city and county, thus making city taxes appear lower than they really are.
The distribution of the services that cities, counties, and states are expected to provide also varies, and is typically a matter of local tradition and political conditions. For example, in one region a city may be responsible for providing services for children, whereas in others this may be a service provided at the county level. A difference in responsibilities means there is also a difference in the governmental distribution of the tax burden, further complicating comparisons.
• Existence of special tax-levying districts. In some regions, special-purpose agencies take on certain service tasks. An example of such an organization is the Bay Area Rapid Transit District (BART), a true government with its own elected officials and taxes. This simply means that it's hard to know what complement of public services is being provided by the taxes collected by a general-purpose government unless you understand the array of special-purpose agencies in that location.
• Disparity in public services. Different types and levels of public services are offered in each city. Cities with extensive public transit networks, large park systems, or high levels of social services naturally have higher costs--and necessarily higher taxes--than cities without these services.
There is no way to fully compensate for all of these factors that complicate the definition of a comparable city, but it is important to understand these limits on the data. This report focuses on the tax burden borne by companies, not on the more typical "good government" questions pertaining to what "you get for the money" paid in taxes. Part III of the report relies on the standardized city tax rates reported by Kosmont Companies, a recognized source of such information. It includes data for both Bay Area jurisdictions and major central cities.
Data on tax rates for the Bay Area, California , and the rest of the U.S. are presented in this section.
Source of Data
The source for these data is the 2003 Kosmont-Rose Institute Cost of Doing Business Survey , a report published by Kosmont Companies (www.kosmont.com) and the Rose Institute of State and Local Government at Claremont McKenna College . This survey looks at tax rates, tax incentives, and development policies in hundreds of California jurisdictions, as well as a handful of notable out-of-state jurisdictions. The survey is used by real estate professionals to evaluate sites for development and by companies looking at locational options. It is also used by jurisdictions to evaluate their tax competitiveness vis-à-vis other jurisdictions.
The Kosmont survey evaluates the tax burden based on taxes such as business taxes, telephone taxes, electric taxes, property taxes, and the like. The survey presents business taxes for a variety of types of firms, as tax rates and the impact of tax rates vary across sectors. It classifies firms as general office, professional office, retail, wholesale, manufacturing, personal service, commercial property, or residential property. The numbers the survey presents represent tax costs calculated on the first $10 million in receipts or the first 100 employees, whichever is applicable. For other types of tax such as property tax and utility tax, it simply presents the average tax rate.
The purpose of the survey is to provide consistent, easily comparable tax data--to compare "apples to apples." The method is to construct an imaginary set of "typical firms" and then calculate what these firms would pay in taxes if they located in a given city.
The assumptions used to model typical firms are:
• For most firm types, payroll costs were assumed to be $4 million. Where necessary, it was asummed that this total includes 10 partners, 40 other professionals, and 50 non-professionals.
• Exceptions : For "Professional Office," the assumed payroll is $7 million. For "Wholesale," payroll is $2 million and there is more than $100,000 in capital invested. For "Residential Property," payroll is assumed to be $2 million.
• For most firm types, square-footage occupancy was assumed to be 20,000.
• Exceptions: For "Wholesale," occupancy is assumed to be 200,000 square feet. For "Commercial Property," occupancy is 500,000 square feet, or 100 units--about $10 million in annual rental receipts. For "Residential Property," rental of 925 units with a total area of 740,000 square feet is assumed, again with annual rental receipts of $10 million.
• Firms were assumed to have additional non-payroll operating costs of $1 million.
• Firms were assumed to have net profits of $1 million.
• The average merchandise value held by firms was assumed to equal one-half of gross receipts.
• All figures represent annual fees, but they do not include one-time costs in the initial tax year, or take into account special taxation zones or tax waivers.
Actual taxes paid by real world firms will vary.
The data are provided here in a format designed and provided by the San Francisco Center for Economic Development. SFCED's spreadsheets were modified and expanded by SPUR.
Format of Data
There are two spreadsheets presented:
• Comparison 1. This spreadsheet addresses business taxes for a variety of different firm types. Costs are calculated based on the first $10 million in receipts or the first 100 employees, whichever is applicable for that tax. The spreadsheet also indicates which types of business taxes are levied for each jurisdiction. The spreadsheet includes the Bay Area, California , and out-of-state cities.
• Comparison 2. This spreadsheet lists taxes rates of other types of taxes that affect businesses: electric tax, telephone tax, cellular phone tax, gas tax, water tax, property tax, sales tax, hotel occupancy tax, parking tax, and the document transfer tax. As these taxes generally represent a smaller portion of a company's tax burden than those in Comparison 1, this spreadsheet is limited to Bay Area cities.
San Francisco Business Taxes - A Disclaimer
When reviewing the comparisons below, it is important to remember that San Francisco offers a small business exemption to the payroll tax: firms with a tax due of under $2,500 do not pay any business tax , except for business registration fees. Over 55,000 businesses, or 85% of all city businesses, pay only these fees.
How Does San Francisco Measure Up?
San Francisco is clearly a high-tax city, but the reality is more complex than this. This section will examine the data in brief, summarizing the most important features.
Except for the extraordinarily high taxes assessed to residential and commercial property ventures in Oakland and Berkeley , San Francisco clearly has the highest business taxes in the Bay Area . San Francisco is also remarkable because it uses a payroll tax, the least-common type of tax among Bay Area communities, and because it has substantial utility taxes where some suburban jurisdictions have none. It also levies a heavy parking tax, although this furthers agreed-upon urban planning objectives in addition to generating revenue.
Competition for jobs from other cities is fierce. Some suburban jurisdictions manage to offer extremely low tax rates, and a few offer very simple flat-rate taxes. For example, Palo Alto has no business tax whatsoever and San Ramon charges a flat $250 for all businesses, a basic license fee. While most suburban jurisdictions have higher taxes than Palo Alto and San Ramon, San Francisco 's business tax burden tends to be about 30 to 40 times that of most suburban jurisdictions in the Bay Area.
On the other hand, San Francisco has one of the lowest property tax rates in the region. Oakland and Berkeley stand out as having substantially higher property tax rates, but most cities in the Bay Area are higher on this tax than San Francisco.
The only out-of-state central cities included in these data with higher business taxes than San Francisco are New York City , Philadelphia , and Washington , D.C. This would make San Francisco the fourth most costly central city in terms of taxes in the United States . Business taxes are about 15% higher in New York and Washington than in San Francisco , and Philadelphia 's taxes are a staggering 65% higher.
The fact that a few higher tax cities exist should not necessarily comfort San Francisco policymakers. Washington and New York have had struggling economies for a number of years. Philadelphia 's situation is even worse. It lost as many as 165,000 jobs, controlled for other trends, due to an increase in payroll taxes of 43% between 1970 and 1977. While San Francisco is generally a more favored city for business than Philadelphia , this job loss is an extreme example whose lessons should be heeded.
By contrast, some central cities seem remarkably competitive. Chicago , for instance, has business taxes that could be considered low even by the standards of Bay Area suburbs.
Although these results are striking, it is critical to remember that tax rates alone do not tell the whole story. They do not indicate the value received for tax money in terms of public services. They do not account for differences between regional economies and sociopolitical boundaries.
They also provide no clues as to causation. Some cities have very low taxes because they have poor economies and are trying to attract new firms. Some have high taxes because they are successful places for business and can get away with it.
San Francisco is an expensive place to do business in terms of total business taxes, and yet it is a attractive milieu for many firms. Taxes are only part of the story.
Ady, Robert M. 1997. Discussion of "The Effects of State and Local Public Services on Economic Development". New England Economic Review: Federal Reserve Bank of Boston , March/April, 77-82.
Alonso, William. 1975. Location Theory. In Regional Policy , edited by J. Friedmann and W. Alonso. Cambridge , MA : MIT Press.
Bartik, Timothy J. 1985. Business-location decisions in the United States : Estimates of the Effects of Unionization, Taxes and Other Characteristics of States. Journal of Business and Economic Statistics 3 (1):14-22.
---. 1991. Who Benefits from State and Local Economic Development Policies?Kalamazoo , MI : W. E. Upjohn Institute for Employment Research.
---. 1992. The Effects of State and Local Taxes on Economic Development: A Review of the Research. Economic Development Quarterly 6 (1):102-110.
---. 1999. Jobs, Productivity, and Local Economic Development: What Implications Does Economic Research Have for the Role of the Government" In Tax Policy in the Real World , edited by J. Slemrod. New York : Cambridge University Press.
Blair, John P., and Robert Premus. 1987. Major Factors in Industrial Location: A Review. Economic Development Quarterly 1 (1):72-85.
Bogart, William Thomas. 1998. The Economies of Cities and Suburbs . New Jersey : Prentice Hall.
Bradbury, Katherine, Anthony Downs, and Kenneth Small. 1982. Urban Decline and the Future of American Cities. Washington , D.C. : Brookings Institute.
Bradbury, Katherine, Yolanda K. Kodrzycki, and Robert Tannenwald. 1997. The Effects of State and Local Public Policies on Economic Development: An Overview. New England Economic Review: Federal Reserve Bank of Boston , March/April, 1-12.
Carlton, Dennis. 1983. The Location and Employment Chocies of New Firms: An Econometric Model with Discrete and Continuous Endogenous Variables. The Review of Economics and Statistics 65:440-449.
Charney , Alberta . 1983. Intraurban Manufacturing Location Decisions and Local Tax Differentials. Journal of Urban Economics 14:184-205.
Ciccone, Antonio, and Robert Hall. 1996. Productivity and the Density of Economic Activity. American Economic Review 86:54-70.
Coy, Peter. 2003. Taxes: What Other Cities Can Learn From New York . BusinessWeek , June 16, 74 .
Crapo, John. July 7, 2003 . Interview.
Crapo, John R. 2001. The Impact of the Payroll Tax in San Francisco . San Francisco : San Francisco Chamber of Commerce.
Dalenberg, Douglas R., and Mark D. Partridge. 1995. The Effects of Taxes, Expenditures, and Public Infrastructure on Metropolitan Area Employment? Journal of Regional Science 35 (4):617-640.
Erickson, Rodney, and Michael Wasylenko. 1980. Firm Location and Site Selection in Suburban Municipalities. Journal of Urban Economics 8:69-85.
Fisher, Peter S., and Alan H. Peters. 1997. Tax Spending Incentives and Enterprise Zones. New England Economic Review: Federal Reserve Bank of Boston , March/April, 109-130.
Fisher, Ronald C. 1997. The Effects of State and Local Public Services on Economic Development. New England Economic Review: Federal Reserve Bank of Boston , March/April, 109-130.
Florida , Richard L. 2002. The Rise of the Creative Class and How it's Transforming Work, Leisure, Community and Everyday Life . New York , NY : Basic Books.
Fox, William. 1981. Fiscal Differentials and Industrial Location: Some Empirical Evidence. Urban Studies 18:105-111.
Fox, William F. 1997. Discussion of "Tax and Spending Incentives and Enterprise Zones". New England Economic Review: Federal Reserve Bank of Boston , March/April, 142-146.
Glaeser, Edward L. 1996. Why Economists Still Like Cities. City Journal 6:70-77.
Glaeser, Edward L. 1998. Are Cities Dying? Journal of Economic Perspectives 12 (2):139-160.
Grieson, Ronald E. 1980. Theoretical analysis and emperical measurements of the effects of the Philadelphia income tax. Journal of Urban Economics 8:123-37.
Grubb, Norton W. 1982. The Flight to the Suburbs of Population and Employment, 1960-1970. Journal of Urban Economics :348-367.
Gruenstein, J. M. L. 1980. Jobs in the city: Can Philadelphia afford to raise taxes? Business Review , May/June, 3-11.
Hammond, Jeff, Hardy Merriman, and Gary Wolff. 1999. Equity and Distributional Issues in the Design of Environmental Tax Reform. San Francisco : Redefining Progress.
Haughwout, Andrew F., Robert P. Inman, Steven Craig, and Thomas Luce. 2000. Local Revenue Hills: A General Equilibrium Specification with Evidence from Four U.S. Cities. Cambridge , MA : National Bureau of Economic Research.
---. 2003. Local Revenue Hills: Evidence from Four U.S. Cities. Cambridge , MA : National Bureau of Economic Research.
Haughwout, Andrew, and Robert P. Inman. 2000. Fiscal Policies in Open Cities with Firms and Households. Cambridge , MA : National Bureau of Economic Research.
Helms, L. Jay. 1985. The Effect of State and Local Taxes on Economic Growth: A Time Series-Cross Section Approach. The Review of Economics and Statistics LXVII:574-582.
Hines, James R., Jr. 1996. Altered States: Taxes and the Location of Foreign Direct Investment in America . The American Economic Review 86 (5):1076-1094.
Inman, Robert P. 1987. Philadelphia 's fiscal management of economic transition. In Local and Fiscal Issues in the Philadelphia Metropolitan Area , edited by J. Thomas F. Luce and A. A. Summers. Philadelphia : University of Pennsylvania Press.
---. 1992. Can Philadelphia Escape Its Fiscal Crisis With Another Tax Increase? Business Review: Federal Reserve Bank of Philadelphia :5-20.
King, Philip G. 2001. San Francisco 's Economic Growth, 1995-2000: The Fiscal Health of the City & Implications for the Future. San Francisco : San Francisco Committee on Jobs.
Kosmont, Larry J. 2003. 2003 Kosmont-Rose Institute Cost of Doing Business Survey . www.kosmont.com
Lantsberg, Alex. 2003. A Literature Review of Research into the Links Between Tax Policy and Economic Development, City & Regional Planning, University of California Berkeley, Berkeley, CA.
Luce, Thomas F., Jr. 1994. Local Taxes, Public Services, and the Intrametropolitan Location of Firms and Households. Public Finance Quarterly 22 (2):139-167.
Luce, Thomas F., Jr., and Anita A. Summers. 1987. Local and Fiscal Issues in the Philadelphia Metropolitan Area . Philadelphia : University of Pennsylvania Press.
Lynch, Robert. 1996. Do State and Local Tax Incentives Work? Washington , DC : Economic Policy Institute.
Margolis, Shana. 2003. Business Tax. San Francisco : Office of the Treasurer and Tax Collector, City and County of San Franccisco .
Mark, Stephen T., Therese J. McGuire, and Leslie E. Papke. 1998. The Effect of Taxes on Economic Development. Washington , DC : District of Colombia Tax Revision Commission.
---. 1998. The Effect of Taxes on Employment and Population in the Washington Area. Washington , DC : District of Colombia Tax Revision Commission.