Donate
Independent, objective, nonpartisan research

R 415JSR

Authors

R 415JSR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3711 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(12) "R_415JSR.pdf" ["wpmf_size"]=> string(7) "2053327" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(70185) "www.ppic.org Parcel Taxes as a Local Revenue Source in California April 2015 Jon Sonstelie with research support from Alyssa Baughman Supported with funding from the Donald Bren Foundation SUMMARY A movement is under way in California to transfer authority from the state government to local governments. Public Safety Realignment and the Local Control Funding For- mula have shifted some of the responsibility for corrections and school funding to local authorities. These initiatives are consistent with public opinion, which favors transfer - ring many obligations from the state to localities. However, a stronger local public sector requires not only more responsibility but also adequate revenue, a concern since the passage of Proposition 13 in 1978. That initiative limited property tax increases and replaced locally determined rates with a statewide rate, thereby constraining local government fnances. In many cases, local governments have responded by turning to another source of revenue—the parcel tax. The parcel tax is a tax on parcels of real property collected as part of a property tax bill. Unlike the property tax, the parcel tax cannot be based on property value. Typically, it is a fat tax that does not vary with the size or characteristics of a parcel. To impose a parcel tax, governments must win support from two-thirds of voters. From 2002 to 2012, California cities, school districts, and special districts placed almost 700 parcel tax proposals on the ballot, of which more than half passed. This report evaluates the parcel tax based on standard principles of taxation, including neutrality, equity, stability, simplicity, transparency, integrity, and growth. It fnds that a well- designed parcel tax can be a worthwhile fscal institution for California. Statewide, aggregate property tax revenue may be adequate. But the state’s formula for allocating that revenue is ISTOCK Parcel Taxes as a Local Revenue Source in California 2 www.ppic.org not aligned with the specifc public services that community residents demand and are will - ing to pay for. The parcel tax can strengthen the link between local governments and com - munity residents, promoting government efciency and realistic expectations about what the local public sector can achieve. Because the parcel tax is essentially a tax on land, it has several advantages. Other taxes, such as the sales and income taxes, can distort economic activity by discouraging work or consumption. But land is immobile and limited in supply. A properly designed tax on land tends to be more neutral in its efect on the economy. To promote equity, a parcel tax should ideally be based on land value. However, Proposition 13 prohibits taxes based on land value. For that reason, it is preferable that parcel taxes not be fat but instead be based on parcel size. A well-designed parcel tax is a levy per square foot applied uniformly to all land uses. Such a tax has one major shortcoming though: a parcel tax based on parcel size can make ownership of large tracts of vacant land uneconomical. A lower tax rate for these parcels may be the solution, but that would violate uniformity. Thus, large tracts of vacant land with little value represent a challenge for the parcel tax. Other states do not levy parcel taxes. California turned to it only because of Proposition 13’s limits on the property tax. Yet in this large and complex state, the parcel tax plays a use - ful role, helping California’s diverse localities tailor public services to the needs and desires of their communities. It is vital then that California make the best possible use of this tax to promote government efciency and help ensure that residents get the services they are will - ing to pay for. For the full report and related resources, please visit our publication page: www.ppic.org/main/publication.asp?i=1145 3 Parcel Taxes as a Local Revenue Source in California www.ppic.org Introduction Since 2011, at the urging of Governor Jerry Brown, the California Legislature has enacted two initiatives to transfer authority from the state government to local governments. Assembly Bills 109 and 117, the Public Safety Realign- ment of 2011, transferred inmates convicted of less serious ofenses from state prisons to county jails. Te Local Con- trol Funding Formula, enacted as part of the 2013–14 state budget, loosened state-imposed restrictions on how school districts can spend their funds. Te impetus for both ini- tiatives is the widespread belief that local governments are more creative, more responsive, and more efcient than the state and federal governments. For example, a 2013 Public Policy Institute of California Statewide Survey found that 71 percent of respondents favored shifing tax dollars and responsibility from state to local government (Baldassare et al. 2013). A stronger local public sector requires not only more responsibility but also revenue to match that responsibility. Tat has been a concern since Proposition 13 was passed in 1978. Before then, local governments in California levied their own property tax rates. Proposition 13 replaced those locally determined rates with a statewide 1 percent rate. Tat was less than half the average tax rate at the time, sharply reducing property tax revenue. However, property tax revenue was considerably higher in California than in other states. Te reduction merely brought California in line with the rest of the country. Overall, considering all types of state and local governments taxes, California’s revenue per capita is still greater than that of other states. While state and local tax revenue may be adequate in the aggregate, it is not allocated among governments accord - ing to current needs. Counties levy the 1 percent property tax rate and allocate the revenue among local governments according to a formula devised by the state legislature. Te formula, based on revenue patterns that are more than 30 years old, is no longer aligned with the public services that community residents demand and are willing to pay for. As a result, many local governments have turned to other sources of revenue. One of these is the parcel tax, a tax on parcels of real property collected as part of a parcel’s property tax bill. Unlike the property tax, the parcel tax may not be based on the value of property. It is typically a fat tax per parcel that does not vary with the size or char- acteristics of the parcel. Parcel taxes must be approved by two-thirds of voters. In the communities that levy it, the parcel tax is more than a revenue source. For a local public sector to be efec- tive, a clear link is needed between the public services residents use and the taxes they pay. If local governments have the authority to set rates for a tax that most residents pay, those residents will be more likely to hold govern- A stronger local public sector requires not only more responsibility but also revenue to match that responsibility. To impose a parcel tax, governments must win support from two-thirds of voters. ISTOCK Parcel Taxes as a Local Revenue Source in California 4 www.ppic.org ments accountable for how that tax revenue is spent. And if governments can only levy taxes that voters approve, voters cannot reasonably fault governments for failing to provide the public with services they are unwilling to pay for. A clear link between services and taxes promotes government efciency and realistic expectations about what govern- ment can achieve. Te property tax played that role before Proposition 13. Te parcel tax may be playing it now. Te parcel tax did not fll that role by design, how- ever. For defenders of Proposition 13, such as Jon Coupal (2013), the parcel tax is a backdoor tactic to circumvent the proposition’s tax limits. Proposition 13 reduced the tax rate on the value of property. But a parcel tax can ofset that reduction with another charge, leaving property tax bills more or less unchanged. Is the parcel tax a sneaky backdoor gambit or a legiti- mate form of local taxation? Other states do not levy parcel taxes, suggesting that it is an inferior method for taxing real property. California turned to it only because of limits on a superior method, the property tax. Is the parcel tax truly an inferior form of taxation? Can it be improved? An examination of these questions is overdue. Te California Tax Foundation has initiated a discussion of the parcel tax with a new study (2014), and this report seeks to contribute to that discussion. Te frst section ofers a summary of the parcel tax’s history in California, tracing its origin to Proposition 13’s limits on property taxes. Te report then examines trends in property tax revenue in the state since Proposition 13 took efect in 1978. Following that, it reviews how cities, school districts, and special dis- tricts use the parcel tax in practice. Te report then evalu- ates the parcel tax according to seven standard principles of taxation, including economic neutrality and equity.Te report concludes with fndings and recommenda- tions. Because the parcel tax is essentially a tax on land, it has several advantages over other taxes. However, since the tax cannot be based on the value of land, it has short- comings. Tese can largely be addressed through proper design, specifcally by basing the tax on a parcel’s size. Such a reform could make the parcel tax an even more use- ful revenue option for California’s local governments. Origins Te parcel tax is a California phenomenon, levied by a number of the state’s cities and school districts. Special districts, typically organized in unincorporated areas, also levy parcel taxes to provide particular public services, such as fre and police protection, parks, and libraries. Tough reliable data on the parcel tax are hard to obtain, about 10 percent of cities and school districts levy a parcel tax. Tax rates are relatively small. Of the parcel taxes enacted by school districts between 2002 and 2012, 86 percent were fat rate—that is, the same regardless of the parcel’s size or use. Te median was $96 per parcel. Among cities that enacted fat-rate parcel taxes during this period, the median was $60 per parcel. For special districts, the median was $68 per parcel. By comparison, assuming a 1 percent tax rate, the average tax bill for a California single-family home was $3,236 in 2013–14 (California State Board of Equalization 2014). Te parcel tax was directly shaped by Proposition 13. Anticipating that the new limits on property tax revenue would cause local governments to increase other taxes, the authors of Proposition 13 included a provision to make those increases more difcult to impose by requiring approval by two-thirds of voters. Section 4 of the proposition stipulates: Cities, Counties and special districts, by a two-thirds vote of the qualifed electors of such district, may impose spe- cial taxes on such district, except ad valorem taxes on real property or a transaction tax or sales tax on the sale of real property within such City, County or special district. Is the parcel tax a sneaky backdoor gambit or a legitimate form of local taxation? 5 Parcel Taxes as a Local Revenue Source in California www.ppic.org Te following year the legislature authorized local governments to levy special taxes on parcels of property to support fre and police services. 1 Similar legislation for school districts, community colleges, and libraries followed in 1987 and 1988. 2 Section 4 was not clear about what taxes are special taxes that require support from two-thirds of voters. In 1980, San Francisco tested the distinction between special and general taxes by declaring that a payroll and gross receipts tax was a general tax, and therefore did not require two-thirds voter support. California’s Supreme Court agreed, defning special taxes to be taxes earmarked for particular purposes. Taxes that are not earmarked are general taxes and do not require a two- thirds vote. 3 Tis ruling quickly led to a second initiative in 1984, Proposition 36, which would have required a two-thirds vote for any tax increase. Te initiative failed. Defenders of Proposition 13 responded with a second initiative, Proposition 62. Te proposition accepted the Supreme Court defnition of special and general taxes, but added a requirement that general taxes receive approval of a majority of voters. Proposition 62 passed, giving local governments a clear path around the two- thirds vote requirement. However, there was no such path for parcel taxes. In 1988, the California Court of Appeals ruled that any gen- eral tax on property must be based on the value of property and thus fall under the Proposition 13 limit. A special tax need not be based on value, however. Terefore, all parcel taxes are special taxes, requiring a two-thirds vote, a ruling codifed by Proposition 218 in 1996. Proposition 218 also placed tighter restrictions on special assessments, another form of taxing parcels of property (Sonstelie 2014). Tis long history of legislation, popular initiatives, and court rulings has led to these consequences: 4 • Parcel taxes are necessarily special taxes that require approval by two-thirds of voters. • In addition to the parcel tax, cities have several sources of discretionary tax revenue, such as the utility users tax, business license tax, and sales tax. Tese are not gener- ally special taxes, and they require the approval of only a majority of voters. • For school districts and special districts, the parcel tax is the only real discretionary source of tax revenue. Parcel taxes have become synonymous with special taxes. In theory, cities could label other types of taxes as special taxes. But that would merely increase the threshold for getting voter approval. Anticipating that the new limits on property tax revenue would cause local governments to increase other taxes, the authors of Proposition 13 included a provision to make those increases more difcult to impose by requiring approval by two-thirds of voters. For school districts and special districts, the parcel tax is the only real discretionary source of tax revenue. AP PHOTO/THE RECORD SEARCHLIGHT, GREG BARNET TE Parcel Taxes as a Local Revenue Source in California 6 www.ppic.org Property Tax Revenue Proposition 13’s limits on property tax revenue set the stage for the parcel tax. Te limits had two parts. Te frst capped the property tax rate at 1 percent. Te second capped the growth of assessed value. From a base deter- mined by the purchase price of a property, assessed value can grow no faster than either the infation rate or 2 per- cent per year—whichever is lower. 5 Property tax revenue declined dramatically afer the passage of Proposition 13 (see fgure 1). In 1977, property tax revenue per capita was 69 percent higher in California than in the rest of the country. Te following year, revenue per capita immediately fell to the level of other states. Since then, it has grown at roughly the same rate as the rest of the country. Immediately afer Proposition 13 passed, the state replaced much of the property tax revenue lost by local gov - ernments with state aid, program shifs, and other forms of assistance. Since then, California has greatly reduced this aid to most local governments. But it continues for schools, funded by high state income and sales tax rates. Overall, considering all types of state and local government taxes, California still has greater revenue per capita than other states. Te gap narrowed afer Proposition 13, but Cali - fornia continues to have relatively high state and local tax revenue (see fgure 2). It is difcult to argue that California governments need more revenue in aggregate. However, the distribution of that revenue is an issue. Proposition 13’s authors deferred to the state legislature regarding how the revenue from the 1 percent property tax rate would be allocated among local governments. Te legislature responded by creating a formula that allocated the revenue from each parcel to local jurisdictions in rough proportion to the revenue received by them from that Figure 2. California’s state and local taxes remain above average despite Proposition 13 SOURCES: U.S. Census Bureau, Annual Survey of State and Local Government Finances, 1957–2011; U.S. Census Bureau, Census of Governments, 1957–2011. 0 2,000 1,000 3,000 4,000 5,000 6,000 Dollars per capita (2011$) All other states California 1957 1961 1965 1969 19771973 1981 1985 200520011997199319892009 Figure 1. California’s property tax revenue fell sharply after Proposition 13 SOURCES: U.S. Census Bureau, Annual Survey of State and Local Government Finances, 1957–2011; U.S. Census Bureau, Census of Governments, 1957–2011. 0 1,200 1,000 800 600 400 200 1,400 1,600 1,800 2,000 Dollars per capita (2011$) All other states California 1957 1961 1965 1969 19771973 1981 1985 200520011997199319892009 Immediately after Proposition 13 passed, the state replaced much of the property tax revenue lost by local governments with state aid, program shifts, and other forms of assistance. Since then, California has greatly reduced this aid to most local governments. 7 Parcel Taxes as a Local Revenue Source in California www.ppic.org parcel before Proposition 13 (Doerr 2000). Te formula has evolved since then, but the initial provisions still guide revenue allocation.Te current allocation of property tax revenue has a number of questionable features (Hill 2000). One is due to the mechanics of the state’s school fnance system. Ironically, the property tax revenue allocated to school districts by the legislature’s formula has signifcant implications for cities and special districts, but almost no implications for school districts themselves. Te state establishes revenue targets for each school district, which it has substantially equal - ized per pupil over time. A district’s target is met by either property tax revenue or state aid. If a school district receives an unusually large allocation of property tax revenue, less is available for the cities, counties, and special districts serving the same area. However, the high level of property tax rev - enue directed to the school district reduces its state aid dol - lar for dollar. Te district itself is no better of than another school district that receives very little property tax revenue. Te chief benefciary is the state treasury. Table 1 shows the diferences across counties in the percentage of property tax revenue allocated to schools. Orange and Alameda Counties have approximately the same property tax revenue per capita, but only 23 percent of that revenue goes to schools in Alameda County com- pared with 41 percent in Orange County. As a result, other local governments in Orange County get about 20 percent less property tax revenue per capita than their counter- parts in Alameda County. Table 1 also shows that property tax revenue per capita varies considerably among counties. In Contra Costa County, the 1 percent rate raised $1,404 per capita in 2009–10 Table 1. Property tax revenue in the ten largest counties, 2009–2010 County Per Capita Property Tax Revenue ($) Percentage of Property Tax Revenue for School Districts Per Capita Property Tax Revenue for Other Local Governments ($) Santa Clara 1,71443974 Contra Costa 1, 4 0 4 31966 Orange 1, 39 6 41827 Alameda 1, 310231, 0 11 San Diego 1, 27437804 Los Angeles 1, 0 95 19887 Riverside 98327717 Sacramento 89027645 San Bernardino 85516717 Fresno 65429465 SOURCES: California State Board of Equalization, California State Board of Equalization 2008–09 Annual Report , 2009; and California Department of Education, Standardized Account Code Structure (SACS) unaudited actuals, 2009–10. NOTE: Property tax revenue is 1 percent of taxable assessed valuation. School revenue is from SACS, 2009–10. Property tax revenue for school districts does not include revenue for county ofces of education or community colleges. Dollar values are rounded to the nearest whole number. Ironically, the property tax revenue allocated to school districts by the legislature’s formula has signifcant implications for cities and special districts, but almost no implications for school districts themselves. Parcel Taxes as a Local Revenue Source in California 8 www.ppic.org compared with only $654 in Fresno County. In both counties, about 30 percent of that revenue goes to schools, leaving only $465 per capita for other local governments in Fresno County, compared with $966 per capita in Contra Costa County. Of course, it is no surprise that the same tax rate raises more revenue in Contra Costa County, where housing prices are relatively high, than in Fresno County, where prices are lower. Te point is that the same rate applied throughout the state has very diferent implications for property tax bills and property tax revenue.Even within the same county, the allocation of prop- erty tax revenue can be very uneven. Table 2 shows per capita property tax revenue of ten cities in Los Angeles County in 2009–10. All ten are full-service cities that provide police and fre protection, libraries, and parks. Te third column lists each city’s percentile place in per capita property tax revenue among all 26 full-service cities in the county. Pomona, in the 11th percentile, receives property tax revenue of $85 per capita, while San Marino, in the 92nd percentile, receives $721 per capita. Te property tax diferences are partly related to median household income. Wealthier households live in more-expensive homes, yield - ing greater property tax revenue. But the distribution of commercial and industrial property is also relevant. Los Angeles and Long Beach are in higher percentiles of per capita property tax revenue than their median household incomes would indicate, largely because of high concentra- tions of commercial and industrial property. Te uneven distribution of property tax revenue is one reason for such discretionary local revenue sources as the parcel tax. Te legislature could improve the allocation of property tax revenue. But it is difcult to imagine that any centralized allocation formula could adequately address varying demands for public services across a state as large and diverse as California. Te demand for public services implies a willingness to pay for those services. It is difcult to assess demand without attaching that demand to an obligation to pay for it. Even if the legislature could devise a better formula, local governments should have a discretionary revenue source, particularly one, such as the parcel tax, that Table 2. Property tax revenue for selected Los Angeles County cities, 2009–2010 City Per Capita Property Tax Revenue ($) Percentile among Cities in Per Capita Property Tax Revenue Median Household Income ($) Pomona 8511 48,864 Alhambra 931953 , 917 Santa Fe Springs 1173154, 551 Arcadia 15 0427 7, 3 4 2 Long Beach 17 85052,900 Burbank 238616 7, 6 9 3 Los Angeles 2436549, 745 Pasadena 262736 8 , 310 Redondo Beach 283809 8 , 816 San Marino 72192139,12 2 SOURCES: California State Controller’s Ofce, Cities Annual Report , 2009–10; and U.S. Census Bureau, 2010 Census. The uneven distribution of property tax revenue is one reason for such discretionary local revenue sources as the parcel tax. 9 Parcel Taxes as a Local Revenue Source in California www.ppic.org requires voter approval. Tax-rate referenda force govern - ments to convince voters that they are spending tax dollars wisely, on valuable public services. Referenda also promote realistic expectations about the services that governments can supply. If voters do not approve a proposed tax, they cannot reasonably expect their governments to provide the services that tax would have fnanced. Using the Parcel Tax Tis section provides details on how cities, school districts, and special districts use the parcel tax. 6 Each of these types of government operates in a diferent fscal environment. Cities have more sources of discretionary revenue than school districts and special districts, and school districts have more state aid than special districts. Nevertheless, these three types of government have had similar experi- ences with parcel taxes. • Te frequency of parcel tax elections is roughly the same among them. From 2002 to 2012, cities placed 124 parcel tax proposals on the ballot. As of 2010, California had 481 cities, so the ratio of proposals to cities was approxi- mately 0.26. For school districts, the ratio was 0.34 and for special districts 0.10. • Te passage rate for parcel tax proposals was similar for each type of government: 48 percent for cities, 60 percent for school districts, and 45 percent for special districts. Local governments do not place parcel tax proposals on the ballot without a reasonable chance of passage. • Most parcel taxes are fat-rate taxes. From 2002 to 2012, fat- rate taxes represented 86 percent of the 389 parcel taxes pro - posed by school districts, 51 percent of those proposed by cities, and 75 percent of those proposed by special districts. • Most fat-rate parcel taxes were relatively small. Te median was $60 per parcel for cities, $96 for school dis- tricts, and $68 for special districts. Cities Cities have several revenue sources. Te most important is the property tax, which varies from $95 per capita in the 20th percentile of full-service cities to $380 per capita in the 80th percentile (see table 3). In California, 110 cities are full service, providing fre and police protection, libraries, and parks. Te state has 371 other cities that ofer some of those services and depend on special districts and counties to provide the others. Te state’s formula allocates property tax to cities. Te most important sources of discretionary revenue are the utility users, business license, and transient lodging taxes. Te frst is a tax on the use of utilities, such as gas, electric- ity, telephone, and cable television. Te business license tax is commonly levied on the gross receipts of businesses. Te lodging tax is levied on people staying temporarily in hotels, motels, or other lodging facilities. Following Propo- sition 218 in 1996, all three taxes required approval by a majority of voters. 7 No state agency collects data on parcel tax revenue for cities, and cities do not always clearly identify such rev- enue in their reports. Te best data on parcel taxes come from the California Local Government Finance Almanac , an online resource maintained by the municipal fnance expert Michael Coleman, which has tracked local tax elections since 2002 (Coleman n.d.). From 2003 through 2012, California cities have placed 124 parcel tax propos- als on the ballot, of which 59 received support from at least two-thirds of voters (see fgure 3). Cities placed an unusu- ally large number of proposals on the ballot in 2004. Ten came from a group of cities in Los Angeles County, each of which proposed a tax of $25.26 per parcel to beneft librar- ies. All these proposals failed. From 2003 through 2012, California cities have placed 124 parcel tax proposals on the ballot, of which 59 received support from at least two-thirds of voters. Parcel Taxes as a Local Revenue Source in California 10 www.ppic.org Cities in the San Francisco Bay Area are more likely to have passed a parcel tax than cities in other areas (see fgure 4). Between 2003 and 2012, 7 of the 11 cities in Marin County and 5 of the 14 cities in Alameda County passed a parcel tax. Seven cities in Los Angeles County passed a parcel tax during this period, but this represents only 8 percent of all cities in the county. Eleven counties had only one city that passed a parcel tax between 2003 and 2012, and 41 counties had no parcel tax passed by a city during this period. Of the 106 proposals for which information on tax rates is readily available, 51 provided for a fat tax applied to all properties. Te lowest was an annual tax of $12 per parcel for animal control in Paradise. Te highest was $950 for public safety in Ross. Twenty-nine proposals had a tax of $30 per parcel or less, and 16 had a tax of $100 per parcel or more. Te median was $60 per parcel. Proposals without an overall fat rate ofen involved a fat rate for residential parcels or dwelling units and a diferent rate for nonresi- dential properties. A few parcel taxes are obvious violations of long- standing principles of taxation. For example, most states require property to be assessed uniformly, regardless of use. A few states allow diferent tax rates for diferent classes of property, but limit the number of classes. By contrast, the city of Huntington Park passed a parcel tax in 2004 with 17 diferent rates, depending on parcel use. Te lowest was $20.24 per year for unimproved lots, and the highest was $629.14 per year for hotels, motels, depart- ment stores, mobile home parks, and supermarkets. Te tax was $78.64 for parcels with one single-family residence, $147.18 for parcels with two dwelling units, and so on. Te parcel tax had its origins in a special assessment for a lighting and landscaping district. Te ordinance proposing the tax also repealed the special assessment that Table 3. General revenue per capita of full-service cities, 2009–2010 20th Percentile ($)Median ($)80th Percentile ($) Property tax 95191 380 Sales and use tax 99147 2 31 Utility users tax 00124 Vehicle license fee 7178 83 Business license tax 818 40 Transient lodging tax 618 87 Franchises 1532 50 Real property transfer tax 24 13 All other taxes 00 12 To t a l 4096301, 037 SOURCE: California State Controller’s Ofce, Cities Annual Report, 2009–10. NOTE: Average revenue per capita is weighted by city population. Figure 3. Parcel tax elections, California cities, 2003–2012 45 40 35 30 25 20 15 10 5 0 Proposals 2009–10 2007–08 2005–06 2003–04 2011–12 FailedPassed SOURCE: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac, 2003–12, www.CaliforniaCityFinance.com. 11 Parcel Taxes as a Local Revenue Source in California www.ppic.org was previously levied by the district. Huntington Park ’s city attorney wrote in an analysis of the proposal that the previous assessments, which were based on street frontage, fell more heavily on single-family residences than on mul- tifamily residences, and that the proposed parcel tax would rectify this. Te proposal won 69 percent of the vote in the November 2004 general election.Te Huntington Park tax rates may have been moti- vated by the beneft principle—that is, the desire to tax parcels in rough proportion to their benefts from public services. In contrast, in Desert Hot Springs, a proposed parcel tax was more likely motivated by a desire to export some of the city’s tax burden to residents of other juris- dictions. Te city frst passed a public safety parcel tax in 2000, with 79 percent approval. It renewed the tax in 2010, with 83 percent approving. For the June 2014 election, the city placed on the ballot a proposal to amend its parcel tax by increasing the tax rate on vacant land from $30 to $372 per acre. Te increase would have had a particularly large impact on a Canadian company that owned 1,600 acres of vacant land in the city. Overall, the city projected that the tax would increase revenue by $3.8 million per year (Ramseth 2014). Te proposal received a positive vote from 61 percent of voters, 5 percentage points short of the required two-thirds. Te Desert Hot Springs proposal illustrates that the revenue that cities raise from these tax rates depends on the tax base. Te median California city has 0.32 parcels per capita (see table 4). Tus, for the average city, a $100-per- parcel tax raises $32 per capita. Eighty-seven percent of parcels are residential, so a fat parcel tax is borne mainly by owners of homes and apartments. Revenue also depends on the tax rate that cities can levy, which is related to the value of parcels. A densely populated city may have relatively few parcels per capita. But land values are higher in cities than in rural areas, and land value is a good measure of the tax rate that a parcel can support. In other words, cities may have fewer parcels per capita than rural areas, but a greater tax capac- ity per parcel. Table 4. Parcel tax base, California cities, 2009–2010 20th Percentile Median 80th Percentile Parcels per capita 0.250. 32 0.42 Residential parcels per capita 0. 21 0.270.34 Residential parcels/ total parcels 0.76 0.87 0.92 SOURCE: DataQuick. NOTE: Averages are weighted by population. Figure 4. Percentage of cities in a county that approved parcel taxes, 2003–2012 SOURCE: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac , 2003–12, www.CaliforniaCityFinance.com. Cities in a county with a successful parcel tax election 0 0 to 25% 25% or more Revenue also depends on the tax rate that cities can levy, which is related to the value of parcels. A densely populated city may have relatively few parcels per capita. But land values are higher in cities than in rural areas, and land value is a good measure of the tax rate that a parcel can support. Parcel Taxes as a Local Revenue Source in California 12 www.ppic.org No single source has data on parcel tax revenue for all cities, but 40 cities posted parcel tax revenue on their websites. Among those cities, median revenue per capita was $33 in 2009–10. In the 20th percentile, revenue was $12 per capita, and in the 80th, $104 per capita. School Districts School districts have fewer revenue sources than cities. In 2009–10, California had 963 school districts, with an average revenue of $8,801 per pupil. On average, districts received $2,210 per pupil in property tax revenue. In most districts, property tax revenue was well short of the targets established by the state. General-purpose state aid made up the diference, averaging $3,114 per pupil. In addition, the state allocated to districts an average of $1,865 per pupil in other forms of aid, such as for special education. School districts also received an average of $1,042 per pupil in federal aid. A variety of other local revenue sources, such as interest, rental income, and interagency transfers, added another $570 per pupil. In this breakdown, parcel tax revenue is included in “other local revenue.” In 2009–10, in all school districts, including those without a parcel tax, parcel tax revenue averaged $53 per student. Tough parcel tax revenue comes from a levy on real property, it does not count as local revenue in calculating state aid. Terefore, an increase in parcel tax revenue does not reduce state aid, in contrast with an increase in property tax revenue. Between 2003 and 2012, school districts placed 329 par - cel tax proposals on the ballot (see fgure 5). Sixty percent of these passed. 8 Most proposals included a sunset provision— usually four to ten years. But many districts renew or increase their parcel taxes when they expire. From 2003 to 2012, 197 proposals passed, of which 83 were renewals or increases. Only 114 districts had one or more successful parcel tax elections during this period. Among those dis- tricts with parcel taxes, revenue per pupil was $213 in the district in the 20th percentile, $577 in the median district, and $1,222 in the 80th percentile. Parcel tax proposals for school districts were con- centrated in the San Francisco Bay Area (see fgure 6). In Marin County, 79 percent of districts passed a parcel tax; in Santa Clara County, 74 percent; in San Mateo County, 70 percent; in Alameda County, 61 percent; and in Con- tra Costa County, 50 percent. Statewide, only 12 percent of school districts passed a parcel tax between 2003 and 2012. Te higher parcel tax frequency in the Bay Area is partly explained by household income (Lang and Sonstelie 2014). Te likelihood of levying a parcel tax is positively related to the average income of households in a district. Bay Area districts tend to have higher household income than districts in the rest of the state. However, that does not account for all the diference between the Bay Area and the rest of the state. Controlling for household income Though parcel tax revenue comes from a levy on real property, it does not count as local revenue in calculating state aid. Therefore, an increase in parcel tax revenue does not reduce state aid, in contrast with an increase in property tax revenue. Figure 5. Parcel tax elections, California school districts, 2003–2012 90 80 70 60 50 40 30 20 10 0 Proposals 2009–10 2007–08 2005–06 2003–04 2011–12 FailedPassed SOURCE: EdSource. 13 Parcel Taxes as a Local Revenue Source in California www.ppic.org and other relevant factors, Bay Area districts are still more likely to levy a parcel tax.Of the 389 parcel taxes proposed by school districts during this period, 86 percent were fat rate. Te average rate was $134 per parcel, and the median was $96. Te tax rate was $195 per parcel in the 80th percentile of propos - als and $59 in the 20th percentile. Te proposals that were not fat rate were mostly based on the square footage of or improvements to the parcel. Several of these proposals had diferent tax rates for residential and nonresidential parcels. A 2008 proposal of that type was challenged in court and overturned by the California Court of Appeals in 2013 ( George J. Borikas v. Alameda Unifed School Dis - trict ). In the decision, the court ruled that taxing diferent classes of property at diferent rates violated legislation that authorized school districts to levy parcel taxes that “apply uniformly to all taxpayers or all real property.” 9 Te California Supreme Court denied a motion to review the decision. Te legislature is now considering a bill that would allow school districts to tax residential and non - residential property at diferent rates. Current statutes do not require other local governments to levy parcel taxes u n i for m l y. Te tax base for fat-rate parcel taxes is unevenly dis- tributed among school districts (see table 5). In 2009–10, the median California unifed K–12 school district had 0.39 parcels per capita. A tax of $100 per parcel would raise $39 per capita in that district. Te 80th percentile had 0.81 parcels per capita. A $100-per-parcel tax would raise $81 per capita. As the second and third rows of the table dem- onstrate, the diference between districts with high and low parcels per capita is due to variations in both residen- tial parcels per capita and the percentage of total parcels that are residential. In terms of school services, pupils, not residents, are relevant. Te fourth row of table 5 shows that the ratio of students to total population varies across unifed school districts. Te net result of these two factors—variations in parcels per capita and students per capita—creates consid- Table 5. Parcel tax base, California unifed school districts, 2009–2010 20th Percentile Median 80th Percentile Parcels per capita 0.290.39 0. 81 Residential parcels per capita 0.23 0.290.38 Residential parcels/ total parcels 0.47 0.780.89 Students per capita 0 .120 .160.20 Parcels per student 1. 6 82 . 416.25 SOURCE: DataQuick. Districts in a county with a successful parcel tax election 0 0 to 25% 25% or more Figure 6. Percentage of school districts in a county that approved parcel taxes, 2003–2012 SOURCE: EdSource.The legislature is now considering a bill that would allow school districts to tax residential and nonresidential property at diferent rates. Parcel Taxes as a Local Revenue Source in California 14 www.ppic.org erable variation in parcels per student. For the median dis- trict in 2009–10, a tax of $100 per parcel would raise $241 per pupil, about 3 percent of the state average of $8,682 per pupil. Te same tax would raise $625 per student in a district in the 80th percentile of parcels per student.Diferences in parcels per student afect more than the yield of a parcel tax. Tey also afect the likelihood that dis - tricts levy a parcel tax. Bree Lang and Jon Sonstelie (2014) fnd that districts with relatively high ratios of parcels to students are more likely to levy parcel taxes than other dis - tricts. However, the ratio of parcels to students is only one factor in the base for a parcel tax. Te value of parcels surely afects the tax rate that a district can levy, and that value ranges widely from densely populated cities to rural areas. Special Districts Special districts are local governments that provide a lim- ited number of services. Examples are fre protection, parks and recreation, and library districts. Special districts have governing boards and taxing authority, but they typically do not have the full range of police powers of cities and counties (Manatt et al. 2010). From 2003 through 2012, special districts placed 238 parcel tax proposals on the ballot (table 6). More than a third of these came from county service areas and commu - nity service districts, mostly for fre protection services and for road construction and maintenance. In addition to the proposals listed in table 6, there was one proposal from an airport district, one from a vector control district, and two from transit districts. Tree of these four proposals passed. Among the 193 proposals from special districts for which tax rate information was available, 142 proposed a fat rate for all parcels, 44 proposed to tax diferent land uses at diferent rates, and 7 proposed a tax on square footage. Te median fat rate was $68 per parcel, $42 at the 20th percentile, and $148 at the 80th percentile. Evaluating the Parcel Tax Tough relatively few local governments currently levy parcel taxes, use of the tax is growing. Is California ready to embrace the parcel tax as a worthwhile fscal institu- Table 6. Parcel tax proposals for special districts, 2003–2012 District Type Number of Districts Proposals Proposals per District Percentage Passed County/community service 1, 20 4860.07 40 Fire protection 363800.22 49 Parks and recreation 10 8240.22 42 Hospital/ambulance/health 84180. 21 72 Library 1480.57 63 Water 16 450.03 20 Memorial 2740 .15 0 Cemetery 25130.01 33 Flood control 5230.06 0 Police protection 331.0 0 67 SOURCES: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac , 2003–12, www.CaliforniaCityFinance.com; and California State Controller’s Ofce, Special Districts Annual Report, 2009–2010 , 2 011. Diferences in parcels per student afect more than the yield of a parcel tax. They also afect the likelihood that districts levy a parcel tax. 15 Parcel Taxes as a Local Revenue Source in California www.ppic.org tion? Te California Tax Foundation (2014) has initiated a discussion of that question, and this report ofers another perspective. Te goal is not to identify changes in the par- cel tax that would make it more popular. A tax that passes costs of a community’s public services on to the residents of other communities may be popular with community residents, but it will not generally be in the interests of the state as a whole.Instead, the goal here is to take a broader view, to step back from the perspective of individual taxpayers and focus on widely accepted principles that apply to many taxes. Tis section focuses on seven standard principles of public fnance: neutrality, equity, stability, simplicity, trans - parency, integrity, and growth. Tese general principles can help guide the discussion of whether the parcel tax is acceptable for the state as a whole and what conditions on its use might make it more acceptable. Such a discussion may lead to statewide legislation restricting the structure of parcel taxes. Te particular circumstances of taxpayers will then determine whether a tax with such a structure has enough local support to win approval. Though relatively few local governments currently levy parcel taxes, use of the tax is growing. From 2003 through 2012, special districts placed 238 parcel tax proposals on the ballot. More than a third of those came from fre districts. ISTOCK Parcel Taxes as a Local Revenue Source in California 16 www.ppic.org Neutrality A tax is neutral if it does not distort market outcomes. Neutrality is ofen referred to as economic efciency. A simple example illustrates the importance of this principle. A family is considering an addition to its home. It would be willing to pay as much as $100,000 for the addition, but the cost is only $90,000. Te family’s surplus is $10,000. However, the addition would increase the assessed value of the family’s home and thus its property taxes now and in the future. If the present value of those property tax increases exceeds $10,000, the family will not build the addition. It loses the $10,000 surplus because of the tax. In public fnance, that loss is called the deadweight loss or excess burden of the tax. Te tax has erased the beneft of a productive activity. In general, a tax will cause a deadweight loss whenever its application diminishes an economically productive activity. An excise tax causes a deadweight loss by reducing purchases of the taxed good. An income tax causes a dead- weight loss by reducing work efort and saving. A property tax causes a deadweight loss by reducing improvements to property. In each case, the size of the loss is determined by the sensitivity of the activity to its price. An income tax has a large deadweight loss if work efort increases signifcantly as wages rise or if saving is very sensitive to interest rates. In choosing taxes to fnance public services, communities should try to minimize deadweight loss. Taxes on land have a special appeal in this regard, because the supply of land is immobile and fxed by nature. As Henry George (1879) argued, a properly administered tax on land has no deadweight loss. To achieve this, the tax must not depend on the use of the land. If the tax is higher on improved land than on unimproved land, it will discourage productive use of land, causing a deadweight loss. Similarly, if the tax is higher on land used for commercial purposes than on land for homes and apartments, it will discourage com - mercial activity, causing a deadweight loss. George argued for a tax on the value of land in its highest and best use. In that case, the tax would not change if land use changed. A parcel tax could have the same benefcial efect as a land value tax (Oates and Schwab 2009). Te key is that the tax must not depend on the use of the land. Tat condition is easily satisfed by a tax on square footage. At frst look, the condition also seems to be satisfed by a fat tax per parcel. However, a fat tax tends to discourage the subdivision of large parcels, which in turn may discour- age residential development. To minimize deadweight loss, a parcel tax should be based on parcel size by levying a determined tax rate per square foot. To avoid deadweight loss, it should not be based on the use of the parcel or characteristics of any developments on the parcel. A parcel tax based on parcel size will have a much lower deadweight loss than an income, sales, or property tax. A uniform tax applied to all parcels in a community also reduces opportunities to export tax burdens to other communities. Te Desert Hot Springs parcel tax proposal is a good example of tax exporting. Te city proposed to increase the tax on vacant land, much of which belonged to residents of other communities, who could not vote on the proposal. Under a uniformity rule, the city could only increase its tax rate on vacant land if residents were also willing to increase the tax on their own parcels by the same amount. Equity Te equity principle is generally separated into two com- ponents: horizontal and vertical. A tax is horizontally equitable if two taxpayers in the same economic position pay the same tax. Te difcult issue is judging whether two taxpayers are in the same position. Income or wealth is commonly used to defne a taxpayer’s position. By that standard, a parcel tax is not horizontally equitable. Two taxpayers with the same income will pay diferent parcel In general, a tax will cause a deadweight loss whenever its application diminishes an economically productive activity. 17 Parcel Taxes as a Local Revenue Source in California www.ppic.org taxes if one owns a home and the other rents. Even if own- ership of land is used to determine a taxpayer’s economic position, a fat tax may not be horizontally equitable. Te main inequity concerns very small parcels created by subdivision. Tese parcels are too small to be developed by themselves and are ofen possessed by the owner of an adjacent parcel who treats the smaller parcel as part of the larger parcel. With a fat tax, the owner would pay a tax on both parcels. By contrast, with a tax on square footage, the total tax on the two parcels would be the same as if the parcels were one. Vertical equity concerns the tax burdens of taxpay- ers in diferent economic positions (Cordes 2011). Vertical equity is frequently judged according to the ability-to-pay principle. Higher-income taxpayers have greater ability to pay and should pay higher taxes, according to this prin- ciple. By that standard, a fat-rate parcel tax is not vertically equitable, because all homeowners pay the same regard- less of their income. A parcel tax based on square foot- age would be better, according to that standard, because wealthier homeowners tend to live on larger parcels. How- ever, the value of a homeowner’s parcel is an even better measure of ability to pay than the size of the parcel. Vertical equity can also be judged according to the beneft principle (Cordes 2011). By that standard, taxes should be related to the beneft that taxpayers receive from public services. A parcel tax looks relatively better accord - ing to this standard. If a community improves public services, it becomes more desirable for potential residents, increasing demand for housing there. If zoning regulations permit, this increased demand will increase the supply of housing. However, over the long run, that increase is limited by the supply of land. So enhancing a community’s public services tends to increase land values. Te increase occurs in both residential and nonresidential areas because nonresidential uses must compete for land with residential uses. Tus, landowners throughout the community are the primary benefciaries of improved public service. Accord- ing to the beneft principle, they should pay for those improvements, which they do if the improvements are fnanced by a parcel tax. Again, a tax on square footage is more equitable than a fat tax because the increase in value from better public services will be roughly proportional to parcel size. Te beneft principle is particularly appropriate for judging the parcel tax because of the way that it is levied. Te tax does not provide most of the funding for local public services. It fnances the marginal improvement of services over a baseline that is fnanced by other revenue. If the total beneft of an enhancement, measured by the willingness of each resident to pay for it, exceeds its cost, there is a strong case for fnancing it. Whether or not it is fnanced depends on voter support. Tat in turn depends on whether a voter’s willingness to pay exceeds the voter’s aversion to the additional tax. Te enhancement will receive voter support if, and only if, total benefts exceed total costs. Of course, perfect proportionality is impossible. But a good alignment of taxes and benefts will tend to make voter support for a proposal more likely when total benefts exceed total costs and less likely when benefts are less than costs. Equity is a difcult principle to apply, not only for the parcel tax but also for sales, income, and property taxes. Tax burdens under the income tax seem most consistent with the ability-to-pay standard. Sales tax and property tax burdens are also related to the ability to pay. By that standard, the property tax seems superior to the parcel tax, because it is not a fat rate per parcel or square foot but a tax based on the value of property. However, when judged according to the beneft principle, the parcel tax seems quite equitable relative to other taxes. A tax on square footage is more equitable than a fat tax because the increase in value from better public services will be roughly proportional to parcel size. Parcel Taxes as a Local Revenue Source in California 18 www.ppic.org Stability Taxpayers should be protected against sudden and unex- pected increases. During the 1970s, California housing prices increased rapidly, pushing up assessed values and property taxes. Tese increases explain at least some of the support for Proposition 13 (Fox 2003). In contrast, the par- cel tax is stable by its very nature, more so than the sales and income taxes. Simplicity Te parcel tax is also simple and less costly to administer than property, sales, and income taxes. In the case of the property tax, each county’s assessor maintains a roll of taxable properties with the names and addresses of parcel owners and various parcel characteristics, including land area. To distribute property tax revenue to local govern - ments, the county property tax administrator has a list of the local governments serving each parcel, which is used to generate property tax bills. Tose bills typically include additional charges, such as sewage and water. A parcel tax is just one more charge added to a parcel’s property tax bill. Transparency Regardless of whether the tax is fat or calculated per square foot, the basis for the parcel tax is objective and easy to understand. Unlike the income tax, there are no long forms to complete or complicated statutes to interpret. And, unlike a sales tax, every taxpayer knows his or her annual tax bill. A parcel tax bill comes once a year as an item listed on a parcel’s property tax bill. Unlike the sales tax, the parcel tax does not involve innumerable, small charges levied throughout the year. Integrity A tax that is easy to evade erodes public confdence in gov - ernment. Like the property tax, parcel taxes are difcult to evade. Parcels are easily identifed, parcel taxes are included on property tax bills, and governments may place a lien on properties with unpaid taxes. Te sales tax is also difcult to evade, but evasion is a problem with the income tax. Growth Tax revenue should grow in line with the cost of the services that the tax fnances. Unlike other taxes, a parcel tax’s base does not grow over time. Tus, it requires tax- rate adjustments to keep pace with rising costs. Some local governments have put in place automatic rate increases tied to the infation rate. Many school districts take a diferent approach. Tax proposals include a sunset provi- sion of usually four to ten years. When the tax expires, the district typically proposes another parcel tax that carries a rate that accounts for infation, enrollment growth, and changes in state aid. Tis approach is more fexible than automatic infation adjustments and also helps build a strong link between the district and its taxpayers. Shortcomings According to the seven principles of taxation, a uniform tax applied to square footage compares favorably with other taxes. However, it has one major shortcoming, and it concerns large parcels of little value. A parcel’s slope may be too steep for development, or the parcel may be far from an urban area and unft for agriculture. In such cases, the stream of future tax payments from a parcel tax may exceed any reasonable expectation of future income. Te application of the parcel tax in this case is akin to confs- cating land, which is unfair by any standard. If the tax were based on land value, this shortcoming could easily be overcome. However, a tax on value would violate the Proposition 13 limit on ad valorem taxes. Another approach would be to tax vacant land at a lower rate. But this approach would penalize development and violate the neutrality principle. Parcels of little value sufer According to the seven principles of taxation, a uniform tax applied to square footage compares favorably with other taxes. 19 Parcel Taxes as a Local Revenue Source in California www.ppic.org an inherent confict between neutrality and equity, which can only be resolved by a tax based on value. Taxing land fairly without taxing the value of land is challenging.Te difculty of this challenge depends on the tax rate. Every property has a tax capacity—that is, the tax rate that would threaten to reduce the property’s value to zero. A low tax rate is unlikely to exceed the capacity of any parcel. How - ever, as the tax rate rises, more properties become vulnerable. Te distribution of tax capacity will also vary by community. In suburban communities, tax capacities are likely to be relatively high across the board, allowing more leeway for tax rates than in more-rural areas, where large portions of land may have little value. Te distribution of tax capacity within communities is an empirical question worth investigating. Conclusion For the authors of Proposition 13, the parcel tax falls into that well-populated category of unintended consequences. Tey surely did not intend to limit property taxes only to create another way to tax real property. Proposition 13 capped property tax revenue, which arguably was excessive. But it also closed of the main avenue through which local governments could address the demands of their constituents. Te state has devised a formula for allocating property tax revenue among local governments. Yet no centralized formula can be expected to respond to the demands of constituents in such a large and diverse state as California. Some form of local, discre- tionary revenue is needed to supplement the property tax revenue allocated to each government. A local, discretion- ary source of revenue is also a crucial link between local governments and their taxpayers, an important factor in promoting an efective local public sector. Te parcel tax has assumed that role in several school districts, special districts, and cities. Tough currently lim - ited, parcel tax use is growing. Tat makes it increasingly important to ask whether it is an appropriate tax to play the role it is playing and what provisions might make it more acceptable. Tis report aims to advance that discussion. Because the parcel tax is a tax on land, it has many advantages over other taxes. Designed properly, it causes virtually no deadweight loss—that is, it does not diminish economic activity the way the sales and income taxes do. Moreover, it assigns taxes in line with benefts, a common standard of equity. Te essence of a properly designed parcel tax is not a fat tax but rather a tax per square foot of land, applied uniformly to all land uses. Such uniformity would also reduce opportunities for communities to export their tax burdens to other communities. Despite these advantages, a uniform tax on square footage has a major shortcoming. Some large parcels may have little value and thus be limited in their capacity to support a parcel tax. Te obvious solution is to base the tax on the value of land. However, a tax on value is prohibited by Proposition 13. Tat restriction is the primary obstacle the parcel tax must overcome. It is a challenge to tax land without taxing the value of the land. Tat challenge has been met in two ways. Te frst is to impose relatively low tax rates that do not exceed the tax capacity of low-value properties. Te Because the parcel tax is a tax on land, it has many advantages over other taxes. Designed properly, it causes virtually no deadweight loss—that is, it does not diminish economic activity the way the sales and income taxes do. The essence of a properly designed parcel tax is not a fat tax but rather a tax per square foot of land, applied uniformly to all land uses. Parcel Taxes as a Local Revenue Source in California 20 www.ppic.org second is to sacrifce uniformity—for example, by levy- ing a lower rate on vacant land. Te issue of uniformity is now before the state legislature, prompted by the Borikas decision, which overturned a parcel tax proposal that levied diferent residential and nonresidential rates. If the legislature insists on uniformity for school districts and extends that principle to cities and special districts, local governments’ use of the parcel tax will grow more slowly. A middle course would be to allow a few specifc deviations from uniformity, such as a lower tax rate for vacant land. Regardless of what the legislature decides, the state should acknowledge the signifcant role the parcel tax plays and begin to monitor it. In particular, the State Controller’s Ofce should collect data on parcel tax rates and revenue, and publish those data in its annual reports on cities, counties, school districts, and special districts. Notes 1 California Government Code, Section 53978. 2 California Government Code, Sections 50079 and 53717. 3 City and County of San Francisco v. Farrell, 32 Cal. 3d 47 (1982). 4 Taylor (2014) provides a fuller description of voter requirements for local taxes. 5 Taylor (2012) provides a fuller description of California’s prop - erty tax system. 6 Much of this material was covered in an earlier and more comprehensive study by Rueben and Cerdan (2003). Because this report focuses on broad-based taxes for general public services, the parcel taxes reviewed in this section are those levied by cities, school districts, and special districts, not the taxes levied by com - munity facility districts created by these other governments. 7 Coleman (2014) provides a detailed description of each revenue source. 8 Chavez and Freedberg (2013) and McGhee and Weston (2013) provide thorough summaries of parcel tax elections for California school districts. 9 California Government Code, Section 50079. 21 Parcel Taxes as a Local Revenue Source in California www.ppic.org References Baldassare, Mark, Dean Bonner, Sonja Petek, and Jui Shrestha. 2013. PPIC Statewide Survey: Californians and Teir Government. San Francisco: Public Policy Institute of California. Available at w w w.ppic.org/main/publication.asp?i=1044. California State Board of Equalization. 2014. A Report on Budgets, Workloads, and Assessment Appeals Activities in Cali- fornia Assessors’ Ofces 2012–2013. Sacramento. Available at www.boe.ca.gov/proptaxes/pdf/lta14003.pdf. California Tax Foundation. 2014. Piecing Together California’s Parcel Taxes: An In-Depth Survey of Local Taxes on Property. Sacramento. Available at w w w.caltaxfoundation.org/reports /parceltax-summary.pdf. Chavez, Lisa, and Louis Freedberg. 2013. Raising Revenues Locally: Parcel Taxes in California School Districts 1983–2012. Oakland, CA: EdSource. Available at edsource.org/wp-content /publications/pub13-ParcelTaxesFinal.pdf. Coleman, Michael. n.d. California Local Government Finance Almanac. Available at w w w.californiacity fnance.com/index .php#VOTES. Coleman, Michael. 2014. Te California Municipal Revenue Sources Handbook. Sacramento: League of California Cities. Cordes, Joseph J. 2011. “Vertical Equity.” In Te Encyclopedia of Taxation and Tax Policy , 2nd ed., edited by Joseph J. Cordes, Robert D. Ebel, and Jane Gravelle. Washington, DC: Urban Institute Press. Coupal, Jon. 2013. “Te Shameful History of Parcel Taxes.” Howard Jarvis Taxpayers Association, February 10. Available at w w w.hjta.org/california-commentary/shameful-history-parcel - t a x e s /. Doerr, David R. 2000. California’s Tax Machine: A History of Taxing and Spending in the Golden State. Sacramento: California Taxpayers Association. Fox, Joel. 2003. Te Legend of Proposition 13: Te Great Califor- nia Tax Revolt. N.p.: Xlibris Corporation. George, Henry. 1879. Progress and Poverty: An Inquiry into the Causes of Industrial Depression and of Increase of Want with Increase in Wealth. New York: Doubleday, Page & Co. Hill, Elizabeth G. 2000. Reconsidering AB 8: Exploring Alterna- tive Ways to Allocate Property Taxes. Sacramento: Legislative Analyst’s Ofce. Available at www.lao.ca.gov/2000/020300 _ab8/020300_ab8.pdf. Lang, Bree, and Jon Sonstelie. 2014. “Te Parcel Tax as a Local Revenue Source for California Public Schools.” Unpublished manuscript, Xavier University and University of California, Santa Barbara. Available at econ.ucsb.edu/~jon/documents /LangSonstelieParcelTax _4_18_14.pdf. Manatt, April, Kimia Mizany, Peter Detwiler, and Elvia Diaz. 2010. What’s So Special about Special Districts? A Citizen’s Guide to Special Districts in California. 4th ed. Sacramento: California Senate Local Government Committee. Available at www.csda .net/wp-content/uploads/2013/04/WhatsSoSpecial_2010.pdf. McGhee, Eric, and Margaret Weston. 2013. Parcel Taxes for Education in California. San Francisco: Public Policy Institute of California. Available at www.ppic.org/content/pubs/atissue /AI_913EMAI.pdf. Oates, Wallace E., and Robert M. Schwab. 2009. “Te Simple Analytics of Land Value Taxation.” In Land Value Taxation: Teory, Evidence, and Practice , edited by Richard F. Dye and Richard W. England. Cambridge, MA: Lincoln Institute of Land Policy. Ramseth, Luke. 2014. “Desert Hot Springs: With Parcel Tax Measure, City Hopes to Avoid Second Bankruptcy.” Riverside Press Enterprise , May 2. Rueben, Kim S., and Pedro Cerdan. 2003. Fiscal Efects of Voter Approval Requirements on Local Governments. San Francisco: Public Policy Institute of California. Available at www.ppic.org /content/pubs/report/R _103KRR.pdf. Sonstelie, Jon. 2014. “California’s Parcel Tax.” Working paper, Lincoln Institute of Land Policy. Available at www.lincolninst .edu/pubs/dl/2460_1808_ SonstelieWP14JS1.pdf. Taylor, Mac. 2012. Understanding California’s Property Taxes. Sacramento: Legislative Analyst’s Ofce. Available at www.lao .ca.gov/reports/2012/ta x/property-ta x-primer-112912.pdf. Taylor, Mac. 2014. A Look at Voter-Approval Requirements for Local Taxes. Sacramento: Legislative Analyst’s Ofce. Available at w w w.lao.ca.gov/reports/2014/fnance/local-taxes /voter-approval-032014.pdf. Parcel Taxes as a Local Revenue Source in California 22 www.ppic.org 22 About the Author Jon Sonstelie is an adjunct fellow and Bren fellow at the Public Policy Institute of California and a professor of economics at the University of California, Santa Barbara. His areas of expertise are public fnance and urban economics, with a focus on school fnance. He coauthored the California School Finance Revenue Manual , which explains the complexities of public school funding in California, and he developed the PPIC School Finance Model for simulating the efects of various school funding formulas. His work includes studies of school budgets and student achievement, state fnance and the Master Plan, school resources and academic standards, and the efect of public school quality on private school enrollment. He was previously a research fellow at Resources for the Future. He holds a PhD in economics from Northwestern University and a BA from Washington State University. Acknowledgments I am grateful to Michael Coleman and EdSource for sharing their data on parcel tax elections. For comments on an earlier draf of this report, I am also grateful to Chas Alamo, Michael Coleman, Rob Gutierrez, Patrick Murphy, Marianne O’Malley, Terri Sexton, Lynette Ubois, Paul Warren, and Joan Youngman. Te Lincoln Institute of Land Policy provided important support in the early stage of this research. Any errors in this work are my own. 23 Parcel Taxes as a Local Revenue Source in California www.ppic.org Board of Directors DONNA LUCAS, CHAIRChief Executive Ofcer Lucas Public Afairs MARK BALDASSAREPresident and CEO Public Policy Institute of California RUBEN BARRALESPresident and CEO GROW Elect MARÍA BLANCOExecutive Director, Undocumented Student Legal Services Center University of California Ofce of the President BRIGITTE BRENAttorney LOUISE HENRY BRYSONChair Emerita, Board of Trustees J. Paul Getty Trust WALTER B. HEWLETTMember, Board of Directors The William and Flora Hewlett Foundation PHIL ISENBERGVice Chair Delta Stewardship Council MAS MASUMOTOAuthor and Farmer STEVEN A. MERKSAMERSenior Partner Nielsen, Merksamer, Parrinello, Gross & Leoni, LLP KIM POLESEChairman ClearStreet, Inc. THOMAS C. SUT TONRetired Chairman and CEO Pacifc Life Insurance Company PPIC is a public charity. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public ofce. PPIC was established in 1994 with an endowment from William R. Hewlett. Copyright © 2015 Public Policy Institute of California. All rights reserved. San Francisco, CA Mark Baldassare is President and Chief Executive Ofcer of PPIC. Donna Lucas is Chair of the Board of Directors. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source. Research publications refect the views of the authors and do not necessarily refect the views of the staf, ofcers, or Board of Directors of the Public Policy Institute of California. Library of Congress Cataloging-in-Publication Data are available for this publication. I S B N 978 -1-5 8213 -16 0 -3 PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 Telephone 415.291.4400 Fa x 415 . 2 91. 4 4 01 PPIC SACRAMENTO CENTER Senator Ofce Building 1121 L Street, Suite 801 Sacramento, California 95814 Telephone 916.440.1120 Fax 916.440.1121 Additional resources related to fscal/governance reform are available at www.ppic.org. The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research." } ["___content":protected]=> string(102) "

R 415JSR

" ["_permalink":protected]=> string(95) "https://www.ppic.org/publication/parcel-taxes-as-a-local-revenue-source-in-california/r_415jsr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8958) ["ID"]=> int(8958) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:42:29" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(4449) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 415JSR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_415jsr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_415JSR.pdf" ["wpmf_size"]=> string(7) "2053327" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(70185) "www.ppic.org Parcel Taxes as a Local Revenue Source in California April 2015 Jon Sonstelie with research support from Alyssa Baughman Supported with funding from the Donald Bren Foundation SUMMARY A movement is under way in California to transfer authority from the state government to local governments. Public Safety Realignment and the Local Control Funding For- mula have shifted some of the responsibility for corrections and school funding to local authorities. These initiatives are consistent with public opinion, which favors transfer - ring many obligations from the state to localities. However, a stronger local public sector requires not only more responsibility but also adequate revenue, a concern since the passage of Proposition 13 in 1978. That initiative limited property tax increases and replaced locally determined rates with a statewide rate, thereby constraining local government fnances. In many cases, local governments have responded by turning to another source of revenue—the parcel tax. The parcel tax is a tax on parcels of real property collected as part of a property tax bill. Unlike the property tax, the parcel tax cannot be based on property value. Typically, it is a fat tax that does not vary with the size or characteristics of a parcel. To impose a parcel tax, governments must win support from two-thirds of voters. From 2002 to 2012, California cities, school districts, and special districts placed almost 700 parcel tax proposals on the ballot, of which more than half passed. This report evaluates the parcel tax based on standard principles of taxation, including neutrality, equity, stability, simplicity, transparency, integrity, and growth. It fnds that a well- designed parcel tax can be a worthwhile fscal institution for California. Statewide, aggregate property tax revenue may be adequate. But the state’s formula for allocating that revenue is ISTOCK Parcel Taxes as a Local Revenue Source in California 2 www.ppic.org not aligned with the specifc public services that community residents demand and are will - ing to pay for. The parcel tax can strengthen the link between local governments and com - munity residents, promoting government efciency and realistic expectations about what the local public sector can achieve. Because the parcel tax is essentially a tax on land, it has several advantages. Other taxes, such as the sales and income taxes, can distort economic activity by discouraging work or consumption. But land is immobile and limited in supply. A properly designed tax on land tends to be more neutral in its efect on the economy. To promote equity, a parcel tax should ideally be based on land value. However, Proposition 13 prohibits taxes based on land value. For that reason, it is preferable that parcel taxes not be fat but instead be based on parcel size. A well-designed parcel tax is a levy per square foot applied uniformly to all land uses. Such a tax has one major shortcoming though: a parcel tax based on parcel size can make ownership of large tracts of vacant land uneconomical. A lower tax rate for these parcels may be the solution, but that would violate uniformity. Thus, large tracts of vacant land with little value represent a challenge for the parcel tax. Other states do not levy parcel taxes. California turned to it only because of Proposition 13’s limits on the property tax. Yet in this large and complex state, the parcel tax plays a use - ful role, helping California’s diverse localities tailor public services to the needs and desires of their communities. It is vital then that California make the best possible use of this tax to promote government efciency and help ensure that residents get the services they are will - ing to pay for. For the full report and related resources, please visit our publication page: www.ppic.org/main/publication.asp?i=1145 3 Parcel Taxes as a Local Revenue Source in California www.ppic.org Introduction Since 2011, at the urging of Governor Jerry Brown, the California Legislature has enacted two initiatives to transfer authority from the state government to local governments. Assembly Bills 109 and 117, the Public Safety Realign- ment of 2011, transferred inmates convicted of less serious ofenses from state prisons to county jails. Te Local Con- trol Funding Formula, enacted as part of the 2013–14 state budget, loosened state-imposed restrictions on how school districts can spend their funds. Te impetus for both ini- tiatives is the widespread belief that local governments are more creative, more responsive, and more efcient than the state and federal governments. For example, a 2013 Public Policy Institute of California Statewide Survey found that 71 percent of respondents favored shifing tax dollars and responsibility from state to local government (Baldassare et al. 2013). A stronger local public sector requires not only more responsibility but also revenue to match that responsibility. Tat has been a concern since Proposition 13 was passed in 1978. Before then, local governments in California levied their own property tax rates. Proposition 13 replaced those locally determined rates with a statewide 1 percent rate. Tat was less than half the average tax rate at the time, sharply reducing property tax revenue. However, property tax revenue was considerably higher in California than in other states. Te reduction merely brought California in line with the rest of the country. Overall, considering all types of state and local governments taxes, California’s revenue per capita is still greater than that of other states. While state and local tax revenue may be adequate in the aggregate, it is not allocated among governments accord - ing to current needs. Counties levy the 1 percent property tax rate and allocate the revenue among local governments according to a formula devised by the state legislature. Te formula, based on revenue patterns that are more than 30 years old, is no longer aligned with the public services that community residents demand and are willing to pay for. As a result, many local governments have turned to other sources of revenue. One of these is the parcel tax, a tax on parcels of real property collected as part of a parcel’s property tax bill. Unlike the property tax, the parcel tax may not be based on the value of property. It is typically a fat tax per parcel that does not vary with the size or char- acteristics of the parcel. Parcel taxes must be approved by two-thirds of voters. In the communities that levy it, the parcel tax is more than a revenue source. For a local public sector to be efec- tive, a clear link is needed between the public services residents use and the taxes they pay. If local governments have the authority to set rates for a tax that most residents pay, those residents will be more likely to hold govern- A stronger local public sector requires not only more responsibility but also revenue to match that responsibility. To impose a parcel tax, governments must win support from two-thirds of voters. ISTOCK Parcel Taxes as a Local Revenue Source in California 4 www.ppic.org ments accountable for how that tax revenue is spent. And if governments can only levy taxes that voters approve, voters cannot reasonably fault governments for failing to provide the public with services they are unwilling to pay for. A clear link between services and taxes promotes government efciency and realistic expectations about what govern- ment can achieve. Te property tax played that role before Proposition 13. Te parcel tax may be playing it now. Te parcel tax did not fll that role by design, how- ever. For defenders of Proposition 13, such as Jon Coupal (2013), the parcel tax is a backdoor tactic to circumvent the proposition’s tax limits. Proposition 13 reduced the tax rate on the value of property. But a parcel tax can ofset that reduction with another charge, leaving property tax bills more or less unchanged. Is the parcel tax a sneaky backdoor gambit or a legiti- mate form of local taxation? Other states do not levy parcel taxes, suggesting that it is an inferior method for taxing real property. California turned to it only because of limits on a superior method, the property tax. Is the parcel tax truly an inferior form of taxation? Can it be improved? An examination of these questions is overdue. Te California Tax Foundation has initiated a discussion of the parcel tax with a new study (2014), and this report seeks to contribute to that discussion. Te frst section ofers a summary of the parcel tax’s history in California, tracing its origin to Proposition 13’s limits on property taxes. Te report then examines trends in property tax revenue in the state since Proposition 13 took efect in 1978. Following that, it reviews how cities, school districts, and special dis- tricts use the parcel tax in practice. Te report then evalu- ates the parcel tax according to seven standard principles of taxation, including economic neutrality and equity.Te report concludes with fndings and recommenda- tions. Because the parcel tax is essentially a tax on land, it has several advantages over other taxes. However, since the tax cannot be based on the value of land, it has short- comings. Tese can largely be addressed through proper design, specifcally by basing the tax on a parcel’s size. Such a reform could make the parcel tax an even more use- ful revenue option for California’s local governments. Origins Te parcel tax is a California phenomenon, levied by a number of the state’s cities and school districts. Special districts, typically organized in unincorporated areas, also levy parcel taxes to provide particular public services, such as fre and police protection, parks, and libraries. Tough reliable data on the parcel tax are hard to obtain, about 10 percent of cities and school districts levy a parcel tax. Tax rates are relatively small. Of the parcel taxes enacted by school districts between 2002 and 2012, 86 percent were fat rate—that is, the same regardless of the parcel’s size or use. Te median was $96 per parcel. Among cities that enacted fat-rate parcel taxes during this period, the median was $60 per parcel. For special districts, the median was $68 per parcel. By comparison, assuming a 1 percent tax rate, the average tax bill for a California single-family home was $3,236 in 2013–14 (California State Board of Equalization 2014). Te parcel tax was directly shaped by Proposition 13. Anticipating that the new limits on property tax revenue would cause local governments to increase other taxes, the authors of Proposition 13 included a provision to make those increases more difcult to impose by requiring approval by two-thirds of voters. Section 4 of the proposition stipulates: Cities, Counties and special districts, by a two-thirds vote of the qualifed electors of such district, may impose spe- cial taxes on such district, except ad valorem taxes on real property or a transaction tax or sales tax on the sale of real property within such City, County or special district. Is the parcel tax a sneaky backdoor gambit or a legitimate form of local taxation? 5 Parcel Taxes as a Local Revenue Source in California www.ppic.org Te following year the legislature authorized local governments to levy special taxes on parcels of property to support fre and police services. 1 Similar legislation for school districts, community colleges, and libraries followed in 1987 and 1988. 2 Section 4 was not clear about what taxes are special taxes that require support from two-thirds of voters. In 1980, San Francisco tested the distinction between special and general taxes by declaring that a payroll and gross receipts tax was a general tax, and therefore did not require two-thirds voter support. California’s Supreme Court agreed, defning special taxes to be taxes earmarked for particular purposes. Taxes that are not earmarked are general taxes and do not require a two- thirds vote. 3 Tis ruling quickly led to a second initiative in 1984, Proposition 36, which would have required a two-thirds vote for any tax increase. Te initiative failed. Defenders of Proposition 13 responded with a second initiative, Proposition 62. Te proposition accepted the Supreme Court defnition of special and general taxes, but added a requirement that general taxes receive approval of a majority of voters. Proposition 62 passed, giving local governments a clear path around the two- thirds vote requirement. However, there was no such path for parcel taxes. In 1988, the California Court of Appeals ruled that any gen- eral tax on property must be based on the value of property and thus fall under the Proposition 13 limit. A special tax need not be based on value, however. Terefore, all parcel taxes are special taxes, requiring a two-thirds vote, a ruling codifed by Proposition 218 in 1996. Proposition 218 also placed tighter restrictions on special assessments, another form of taxing parcels of property (Sonstelie 2014). Tis long history of legislation, popular initiatives, and court rulings has led to these consequences: 4 • Parcel taxes are necessarily special taxes that require approval by two-thirds of voters. • In addition to the parcel tax, cities have several sources of discretionary tax revenue, such as the utility users tax, business license tax, and sales tax. Tese are not gener- ally special taxes, and they require the approval of only a majority of voters. • For school districts and special districts, the parcel tax is the only real discretionary source of tax revenue. Parcel taxes have become synonymous with special taxes. In theory, cities could label other types of taxes as special taxes. But that would merely increase the threshold for getting voter approval. Anticipating that the new limits on property tax revenue would cause local governments to increase other taxes, the authors of Proposition 13 included a provision to make those increases more difcult to impose by requiring approval by two-thirds of voters. For school districts and special districts, the parcel tax is the only real discretionary source of tax revenue. AP PHOTO/THE RECORD SEARCHLIGHT, GREG BARNET TE Parcel Taxes as a Local Revenue Source in California 6 www.ppic.org Property Tax Revenue Proposition 13’s limits on property tax revenue set the stage for the parcel tax. Te limits had two parts. Te frst capped the property tax rate at 1 percent. Te second capped the growth of assessed value. From a base deter- mined by the purchase price of a property, assessed value can grow no faster than either the infation rate or 2 per- cent per year—whichever is lower. 5 Property tax revenue declined dramatically afer the passage of Proposition 13 (see fgure 1). In 1977, property tax revenue per capita was 69 percent higher in California than in the rest of the country. Te following year, revenue per capita immediately fell to the level of other states. Since then, it has grown at roughly the same rate as the rest of the country. Immediately afer Proposition 13 passed, the state replaced much of the property tax revenue lost by local gov - ernments with state aid, program shifs, and other forms of assistance. Since then, California has greatly reduced this aid to most local governments. But it continues for schools, funded by high state income and sales tax rates. Overall, considering all types of state and local government taxes, California still has greater revenue per capita than other states. Te gap narrowed afer Proposition 13, but Cali - fornia continues to have relatively high state and local tax revenue (see fgure 2). It is difcult to argue that California governments need more revenue in aggregate. However, the distribution of that revenue is an issue. Proposition 13’s authors deferred to the state legislature regarding how the revenue from the 1 percent property tax rate would be allocated among local governments. Te legislature responded by creating a formula that allocated the revenue from each parcel to local jurisdictions in rough proportion to the revenue received by them from that Figure 2. California’s state and local taxes remain above average despite Proposition 13 SOURCES: U.S. Census Bureau, Annual Survey of State and Local Government Finances, 1957–2011; U.S. Census Bureau, Census of Governments, 1957–2011. 0 2,000 1,000 3,000 4,000 5,000 6,000 Dollars per capita (2011$) All other states California 1957 1961 1965 1969 19771973 1981 1985 200520011997199319892009 Figure 1. California’s property tax revenue fell sharply after Proposition 13 SOURCES: U.S. Census Bureau, Annual Survey of State and Local Government Finances, 1957–2011; U.S. Census Bureau, Census of Governments, 1957–2011. 0 1,200 1,000 800 600 400 200 1,400 1,600 1,800 2,000 Dollars per capita (2011$) All other states California 1957 1961 1965 1969 19771973 1981 1985 200520011997199319892009 Immediately after Proposition 13 passed, the state replaced much of the property tax revenue lost by local governments with state aid, program shifts, and other forms of assistance. Since then, California has greatly reduced this aid to most local governments. 7 Parcel Taxes as a Local Revenue Source in California www.ppic.org parcel before Proposition 13 (Doerr 2000). Te formula has evolved since then, but the initial provisions still guide revenue allocation.Te current allocation of property tax revenue has a number of questionable features (Hill 2000). One is due to the mechanics of the state’s school fnance system. Ironically, the property tax revenue allocated to school districts by the legislature’s formula has signifcant implications for cities and special districts, but almost no implications for school districts themselves. Te state establishes revenue targets for each school district, which it has substantially equal - ized per pupil over time. A district’s target is met by either property tax revenue or state aid. If a school district receives an unusually large allocation of property tax revenue, less is available for the cities, counties, and special districts serving the same area. However, the high level of property tax rev - enue directed to the school district reduces its state aid dol - lar for dollar. Te district itself is no better of than another school district that receives very little property tax revenue. Te chief benefciary is the state treasury. Table 1 shows the diferences across counties in the percentage of property tax revenue allocated to schools. Orange and Alameda Counties have approximately the same property tax revenue per capita, but only 23 percent of that revenue goes to schools in Alameda County com- pared with 41 percent in Orange County. As a result, other local governments in Orange County get about 20 percent less property tax revenue per capita than their counter- parts in Alameda County. Table 1 also shows that property tax revenue per capita varies considerably among counties. In Contra Costa County, the 1 percent rate raised $1,404 per capita in 2009–10 Table 1. Property tax revenue in the ten largest counties, 2009–2010 County Per Capita Property Tax Revenue ($) Percentage of Property Tax Revenue for School Districts Per Capita Property Tax Revenue for Other Local Governments ($) Santa Clara 1,71443974 Contra Costa 1, 4 0 4 31966 Orange 1, 39 6 41827 Alameda 1, 310231, 0 11 San Diego 1, 27437804 Los Angeles 1, 0 95 19887 Riverside 98327717 Sacramento 89027645 San Bernardino 85516717 Fresno 65429465 SOURCES: California State Board of Equalization, California State Board of Equalization 2008–09 Annual Report , 2009; and California Department of Education, Standardized Account Code Structure (SACS) unaudited actuals, 2009–10. NOTE: Property tax revenue is 1 percent of taxable assessed valuation. School revenue is from SACS, 2009–10. Property tax revenue for school districts does not include revenue for county ofces of education or community colleges. Dollar values are rounded to the nearest whole number. Ironically, the property tax revenue allocated to school districts by the legislature’s formula has signifcant implications for cities and special districts, but almost no implications for school districts themselves. Parcel Taxes as a Local Revenue Source in California 8 www.ppic.org compared with only $654 in Fresno County. In both counties, about 30 percent of that revenue goes to schools, leaving only $465 per capita for other local governments in Fresno County, compared with $966 per capita in Contra Costa County. Of course, it is no surprise that the same tax rate raises more revenue in Contra Costa County, where housing prices are relatively high, than in Fresno County, where prices are lower. Te point is that the same rate applied throughout the state has very diferent implications for property tax bills and property tax revenue.Even within the same county, the allocation of prop- erty tax revenue can be very uneven. Table 2 shows per capita property tax revenue of ten cities in Los Angeles County in 2009–10. All ten are full-service cities that provide police and fre protection, libraries, and parks. Te third column lists each city’s percentile place in per capita property tax revenue among all 26 full-service cities in the county. Pomona, in the 11th percentile, receives property tax revenue of $85 per capita, while San Marino, in the 92nd percentile, receives $721 per capita. Te property tax diferences are partly related to median household income. Wealthier households live in more-expensive homes, yield - ing greater property tax revenue. But the distribution of commercial and industrial property is also relevant. Los Angeles and Long Beach are in higher percentiles of per capita property tax revenue than their median household incomes would indicate, largely because of high concentra- tions of commercial and industrial property. Te uneven distribution of property tax revenue is one reason for such discretionary local revenue sources as the parcel tax. Te legislature could improve the allocation of property tax revenue. But it is difcult to imagine that any centralized allocation formula could adequately address varying demands for public services across a state as large and diverse as California. Te demand for public services implies a willingness to pay for those services. It is difcult to assess demand without attaching that demand to an obligation to pay for it. Even if the legislature could devise a better formula, local governments should have a discretionary revenue source, particularly one, such as the parcel tax, that Table 2. Property tax revenue for selected Los Angeles County cities, 2009–2010 City Per Capita Property Tax Revenue ($) Percentile among Cities in Per Capita Property Tax Revenue Median Household Income ($) Pomona 8511 48,864 Alhambra 931953 , 917 Santa Fe Springs 1173154, 551 Arcadia 15 0427 7, 3 4 2 Long Beach 17 85052,900 Burbank 238616 7, 6 9 3 Los Angeles 2436549, 745 Pasadena 262736 8 , 310 Redondo Beach 283809 8 , 816 San Marino 72192139,12 2 SOURCES: California State Controller’s Ofce, Cities Annual Report , 2009–10; and U.S. Census Bureau, 2010 Census. The uneven distribution of property tax revenue is one reason for such discretionary local revenue sources as the parcel tax. 9 Parcel Taxes as a Local Revenue Source in California www.ppic.org requires voter approval. Tax-rate referenda force govern - ments to convince voters that they are spending tax dollars wisely, on valuable public services. Referenda also promote realistic expectations about the services that governments can supply. If voters do not approve a proposed tax, they cannot reasonably expect their governments to provide the services that tax would have fnanced. Using the Parcel Tax Tis section provides details on how cities, school districts, and special districts use the parcel tax. 6 Each of these types of government operates in a diferent fscal environment. Cities have more sources of discretionary revenue than school districts and special districts, and school districts have more state aid than special districts. Nevertheless, these three types of government have had similar experi- ences with parcel taxes. • Te frequency of parcel tax elections is roughly the same among them. From 2002 to 2012, cities placed 124 parcel tax proposals on the ballot. As of 2010, California had 481 cities, so the ratio of proposals to cities was approxi- mately 0.26. For school districts, the ratio was 0.34 and for special districts 0.10. • Te passage rate for parcel tax proposals was similar for each type of government: 48 percent for cities, 60 percent for school districts, and 45 percent for special districts. Local governments do not place parcel tax proposals on the ballot without a reasonable chance of passage. • Most parcel taxes are fat-rate taxes. From 2002 to 2012, fat- rate taxes represented 86 percent of the 389 parcel taxes pro - posed by school districts, 51 percent of those proposed by cities, and 75 percent of those proposed by special districts. • Most fat-rate parcel taxes were relatively small. Te median was $60 per parcel for cities, $96 for school dis- tricts, and $68 for special districts. Cities Cities have several revenue sources. Te most important is the property tax, which varies from $95 per capita in the 20th percentile of full-service cities to $380 per capita in the 80th percentile (see table 3). In California, 110 cities are full service, providing fre and police protection, libraries, and parks. Te state has 371 other cities that ofer some of those services and depend on special districts and counties to provide the others. Te state’s formula allocates property tax to cities. Te most important sources of discretionary revenue are the utility users, business license, and transient lodging taxes. Te frst is a tax on the use of utilities, such as gas, electric- ity, telephone, and cable television. Te business license tax is commonly levied on the gross receipts of businesses. Te lodging tax is levied on people staying temporarily in hotels, motels, or other lodging facilities. Following Propo- sition 218 in 1996, all three taxes required approval by a majority of voters. 7 No state agency collects data on parcel tax revenue for cities, and cities do not always clearly identify such rev- enue in their reports. Te best data on parcel taxes come from the California Local Government Finance Almanac , an online resource maintained by the municipal fnance expert Michael Coleman, which has tracked local tax elections since 2002 (Coleman n.d.). From 2003 through 2012, California cities have placed 124 parcel tax propos- als on the ballot, of which 59 received support from at least two-thirds of voters (see fgure 3). Cities placed an unusu- ally large number of proposals on the ballot in 2004. Ten came from a group of cities in Los Angeles County, each of which proposed a tax of $25.26 per parcel to beneft librar- ies. All these proposals failed. From 2003 through 2012, California cities have placed 124 parcel tax proposals on the ballot, of which 59 received support from at least two-thirds of voters. Parcel Taxes as a Local Revenue Source in California 10 www.ppic.org Cities in the San Francisco Bay Area are more likely to have passed a parcel tax than cities in other areas (see fgure 4). Between 2003 and 2012, 7 of the 11 cities in Marin County and 5 of the 14 cities in Alameda County passed a parcel tax. Seven cities in Los Angeles County passed a parcel tax during this period, but this represents only 8 percent of all cities in the county. Eleven counties had only one city that passed a parcel tax between 2003 and 2012, and 41 counties had no parcel tax passed by a city during this period. Of the 106 proposals for which information on tax rates is readily available, 51 provided for a fat tax applied to all properties. Te lowest was an annual tax of $12 per parcel for animal control in Paradise. Te highest was $950 for public safety in Ross. Twenty-nine proposals had a tax of $30 per parcel or less, and 16 had a tax of $100 per parcel or more. Te median was $60 per parcel. Proposals without an overall fat rate ofen involved a fat rate for residential parcels or dwelling units and a diferent rate for nonresi- dential properties. A few parcel taxes are obvious violations of long- standing principles of taxation. For example, most states require property to be assessed uniformly, regardless of use. A few states allow diferent tax rates for diferent classes of property, but limit the number of classes. By contrast, the city of Huntington Park passed a parcel tax in 2004 with 17 diferent rates, depending on parcel use. Te lowest was $20.24 per year for unimproved lots, and the highest was $629.14 per year for hotels, motels, depart- ment stores, mobile home parks, and supermarkets. Te tax was $78.64 for parcels with one single-family residence, $147.18 for parcels with two dwelling units, and so on. Te parcel tax had its origins in a special assessment for a lighting and landscaping district. Te ordinance proposing the tax also repealed the special assessment that Table 3. General revenue per capita of full-service cities, 2009–2010 20th Percentile ($)Median ($)80th Percentile ($) Property tax 95191 380 Sales and use tax 99147 2 31 Utility users tax 00124 Vehicle license fee 7178 83 Business license tax 818 40 Transient lodging tax 618 87 Franchises 1532 50 Real property transfer tax 24 13 All other taxes 00 12 To t a l 4096301, 037 SOURCE: California State Controller’s Ofce, Cities Annual Report, 2009–10. NOTE: Average revenue per capita is weighted by city population. Figure 3. Parcel tax elections, California cities, 2003–2012 45 40 35 30 25 20 15 10 5 0 Proposals 2009–10 2007–08 2005–06 2003–04 2011–12 FailedPassed SOURCE: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac, 2003–12, www.CaliforniaCityFinance.com. 11 Parcel Taxes as a Local Revenue Source in California www.ppic.org was previously levied by the district. Huntington Park ’s city attorney wrote in an analysis of the proposal that the previous assessments, which were based on street frontage, fell more heavily on single-family residences than on mul- tifamily residences, and that the proposed parcel tax would rectify this. Te proposal won 69 percent of the vote in the November 2004 general election.Te Huntington Park tax rates may have been moti- vated by the beneft principle—that is, the desire to tax parcels in rough proportion to their benefts from public services. In contrast, in Desert Hot Springs, a proposed parcel tax was more likely motivated by a desire to export some of the city’s tax burden to residents of other juris- dictions. Te city frst passed a public safety parcel tax in 2000, with 79 percent approval. It renewed the tax in 2010, with 83 percent approving. For the June 2014 election, the city placed on the ballot a proposal to amend its parcel tax by increasing the tax rate on vacant land from $30 to $372 per acre. Te increase would have had a particularly large impact on a Canadian company that owned 1,600 acres of vacant land in the city. Overall, the city projected that the tax would increase revenue by $3.8 million per year (Ramseth 2014). Te proposal received a positive vote from 61 percent of voters, 5 percentage points short of the required two-thirds. Te Desert Hot Springs proposal illustrates that the revenue that cities raise from these tax rates depends on the tax base. Te median California city has 0.32 parcels per capita (see table 4). Tus, for the average city, a $100-per- parcel tax raises $32 per capita. Eighty-seven percent of parcels are residential, so a fat parcel tax is borne mainly by owners of homes and apartments. Revenue also depends on the tax rate that cities can levy, which is related to the value of parcels. A densely populated city may have relatively few parcels per capita. But land values are higher in cities than in rural areas, and land value is a good measure of the tax rate that a parcel can support. In other words, cities may have fewer parcels per capita than rural areas, but a greater tax capac- ity per parcel. Table 4. Parcel tax base, California cities, 2009–2010 20th Percentile Median 80th Percentile Parcels per capita 0.250. 32 0.42 Residential parcels per capita 0. 21 0.270.34 Residential parcels/ total parcels 0.76 0.87 0.92 SOURCE: DataQuick. NOTE: Averages are weighted by population. Figure 4. Percentage of cities in a county that approved parcel taxes, 2003–2012 SOURCE: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac , 2003–12, www.CaliforniaCityFinance.com. Cities in a county with a successful parcel tax election 0 0 to 25% 25% or more Revenue also depends on the tax rate that cities can levy, which is related to the value of parcels. A densely populated city may have relatively few parcels per capita. But land values are higher in cities than in rural areas, and land value is a good measure of the tax rate that a parcel can support. Parcel Taxes as a Local Revenue Source in California 12 www.ppic.org No single source has data on parcel tax revenue for all cities, but 40 cities posted parcel tax revenue on their websites. Among those cities, median revenue per capita was $33 in 2009–10. In the 20th percentile, revenue was $12 per capita, and in the 80th, $104 per capita. School Districts School districts have fewer revenue sources than cities. In 2009–10, California had 963 school districts, with an average revenue of $8,801 per pupil. On average, districts received $2,210 per pupil in property tax revenue. In most districts, property tax revenue was well short of the targets established by the state. General-purpose state aid made up the diference, averaging $3,114 per pupil. In addition, the state allocated to districts an average of $1,865 per pupil in other forms of aid, such as for special education. School districts also received an average of $1,042 per pupil in federal aid. A variety of other local revenue sources, such as interest, rental income, and interagency transfers, added another $570 per pupil. In this breakdown, parcel tax revenue is included in “other local revenue.” In 2009–10, in all school districts, including those without a parcel tax, parcel tax revenue averaged $53 per student. Tough parcel tax revenue comes from a levy on real property, it does not count as local revenue in calculating state aid. Terefore, an increase in parcel tax revenue does not reduce state aid, in contrast with an increase in property tax revenue. Between 2003 and 2012, school districts placed 329 par - cel tax proposals on the ballot (see fgure 5). Sixty percent of these passed. 8 Most proposals included a sunset provision— usually four to ten years. But many districts renew or increase their parcel taxes when they expire. From 2003 to 2012, 197 proposals passed, of which 83 were renewals or increases. Only 114 districts had one or more successful parcel tax elections during this period. Among those dis- tricts with parcel taxes, revenue per pupil was $213 in the district in the 20th percentile, $577 in the median district, and $1,222 in the 80th percentile. Parcel tax proposals for school districts were con- centrated in the San Francisco Bay Area (see fgure 6). In Marin County, 79 percent of districts passed a parcel tax; in Santa Clara County, 74 percent; in San Mateo County, 70 percent; in Alameda County, 61 percent; and in Con- tra Costa County, 50 percent. Statewide, only 12 percent of school districts passed a parcel tax between 2003 and 2012. Te higher parcel tax frequency in the Bay Area is partly explained by household income (Lang and Sonstelie 2014). Te likelihood of levying a parcel tax is positively related to the average income of households in a district. Bay Area districts tend to have higher household income than districts in the rest of the state. However, that does not account for all the diference between the Bay Area and the rest of the state. Controlling for household income Though parcel tax revenue comes from a levy on real property, it does not count as local revenue in calculating state aid. Therefore, an increase in parcel tax revenue does not reduce state aid, in contrast with an increase in property tax revenue. Figure 5. Parcel tax elections, California school districts, 2003–2012 90 80 70 60 50 40 30 20 10 0 Proposals 2009–10 2007–08 2005–06 2003–04 2011–12 FailedPassed SOURCE: EdSource. 13 Parcel Taxes as a Local Revenue Source in California www.ppic.org and other relevant factors, Bay Area districts are still more likely to levy a parcel tax.Of the 389 parcel taxes proposed by school districts during this period, 86 percent were fat rate. Te average rate was $134 per parcel, and the median was $96. Te tax rate was $195 per parcel in the 80th percentile of propos - als and $59 in the 20th percentile. Te proposals that were not fat rate were mostly based on the square footage of or improvements to the parcel. Several of these proposals had diferent tax rates for residential and nonresidential parcels. A 2008 proposal of that type was challenged in court and overturned by the California Court of Appeals in 2013 ( George J. Borikas v. Alameda Unifed School Dis - trict ). In the decision, the court ruled that taxing diferent classes of property at diferent rates violated legislation that authorized school districts to levy parcel taxes that “apply uniformly to all taxpayers or all real property.” 9 Te California Supreme Court denied a motion to review the decision. Te legislature is now considering a bill that would allow school districts to tax residential and non - residential property at diferent rates. Current statutes do not require other local governments to levy parcel taxes u n i for m l y. Te tax base for fat-rate parcel taxes is unevenly dis- tributed among school districts (see table 5). In 2009–10, the median California unifed K–12 school district had 0.39 parcels per capita. A tax of $100 per parcel would raise $39 per capita in that district. Te 80th percentile had 0.81 parcels per capita. A $100-per-parcel tax would raise $81 per capita. As the second and third rows of the table dem- onstrate, the diference between districts with high and low parcels per capita is due to variations in both residen- tial parcels per capita and the percentage of total parcels that are residential. In terms of school services, pupils, not residents, are relevant. Te fourth row of table 5 shows that the ratio of students to total population varies across unifed school districts. Te net result of these two factors—variations in parcels per capita and students per capita—creates consid- Table 5. Parcel tax base, California unifed school districts, 2009–2010 20th Percentile Median 80th Percentile Parcels per capita 0.290.39 0. 81 Residential parcels per capita 0.23 0.290.38 Residential parcels/ total parcels 0.47 0.780.89 Students per capita 0 .120 .160.20 Parcels per student 1. 6 82 . 416.25 SOURCE: DataQuick. Districts in a county with a successful parcel tax election 0 0 to 25% 25% or more Figure 6. Percentage of school districts in a county that approved parcel taxes, 2003–2012 SOURCE: EdSource.The legislature is now considering a bill that would allow school districts to tax residential and nonresidential property at diferent rates. Parcel Taxes as a Local Revenue Source in California 14 www.ppic.org erable variation in parcels per student. For the median dis- trict in 2009–10, a tax of $100 per parcel would raise $241 per pupil, about 3 percent of the state average of $8,682 per pupil. Te same tax would raise $625 per student in a district in the 80th percentile of parcels per student.Diferences in parcels per student afect more than the yield of a parcel tax. Tey also afect the likelihood that dis - tricts levy a parcel tax. Bree Lang and Jon Sonstelie (2014) fnd that districts with relatively high ratios of parcels to students are more likely to levy parcel taxes than other dis - tricts. However, the ratio of parcels to students is only one factor in the base for a parcel tax. Te value of parcels surely afects the tax rate that a district can levy, and that value ranges widely from densely populated cities to rural areas. Special Districts Special districts are local governments that provide a lim- ited number of services. Examples are fre protection, parks and recreation, and library districts. Special districts have governing boards and taxing authority, but they typically do not have the full range of police powers of cities and counties (Manatt et al. 2010). From 2003 through 2012, special districts placed 238 parcel tax proposals on the ballot (table 6). More than a third of these came from county service areas and commu - nity service districts, mostly for fre protection services and for road construction and maintenance. In addition to the proposals listed in table 6, there was one proposal from an airport district, one from a vector control district, and two from transit districts. Tree of these four proposals passed. Among the 193 proposals from special districts for which tax rate information was available, 142 proposed a fat rate for all parcels, 44 proposed to tax diferent land uses at diferent rates, and 7 proposed a tax on square footage. Te median fat rate was $68 per parcel, $42 at the 20th percentile, and $148 at the 80th percentile. Evaluating the Parcel Tax Tough relatively few local governments currently levy parcel taxes, use of the tax is growing. Is California ready to embrace the parcel tax as a worthwhile fscal institu- Table 6. Parcel tax proposals for special districts, 2003–2012 District Type Number of Districts Proposals Proposals per District Percentage Passed County/community service 1, 20 4860.07 40 Fire protection 363800.22 49 Parks and recreation 10 8240.22 42 Hospital/ambulance/health 84180. 21 72 Library 1480.57 63 Water 16 450.03 20 Memorial 2740 .15 0 Cemetery 25130.01 33 Flood control 5230.06 0 Police protection 331.0 0 67 SOURCES: Michael Coleman, “Votes on Local Taxes and Other Revenue Measures,” California Local Government Finance Almanac , 2003–12, www.CaliforniaCityFinance.com; and California State Controller’s Ofce, Special Districts Annual Report, 2009–2010 , 2 011. Diferences in parcels per student afect more than the yield of a parcel tax. They also afect the likelihood that districts levy a parcel tax. 15 Parcel Taxes as a Local Revenue Source in California www.ppic.org tion? Te California Tax Foundation (2014) has initiated a discussion of that question, and this report ofers another perspective. Te goal is not to identify changes in the par- cel tax that would make it more popular. A tax that passes costs of a community’s public services on to the residents of other communities may be popular with community residents, but it will not generally be in the interests of the state as a whole.Instead, the goal here is to take a broader view, to step back from the perspective of individual taxpayers and focus on widely accepted principles that apply to many taxes. Tis section focuses on seven standard principles of public fnance: neutrality, equity, stability, simplicity, trans - parency, integrity, and growth. Tese general principles can help guide the discussion of whether the parcel tax is acceptable for the state as a whole and what conditions on its use might make it more acceptable. Such a discussion may lead to statewide legislation restricting the structure of parcel taxes. Te particular circumstances of taxpayers will then determine whether a tax with such a structure has enough local support to win approval. Though relatively few local governments currently levy parcel taxes, use of the tax is growing. From 2003 through 2012, special districts placed 238 parcel tax proposals on the ballot. More than a third of those came from fre districts. ISTOCK Parcel Taxes as a Local Revenue Source in California 16 www.ppic.org Neutrality A tax is neutral if it does not distort market outcomes. Neutrality is ofen referred to as economic efciency. A simple example illustrates the importance of this principle. A family is considering an addition to its home. It would be willing to pay as much as $100,000 for the addition, but the cost is only $90,000. Te family’s surplus is $10,000. However, the addition would increase the assessed value of the family’s home and thus its property taxes now and in the future. If the present value of those property tax increases exceeds $10,000, the family will not build the addition. It loses the $10,000 surplus because of the tax. In public fnance, that loss is called the deadweight loss or excess burden of the tax. Te tax has erased the beneft of a productive activity. In general, a tax will cause a deadweight loss whenever its application diminishes an economically productive activity. An excise tax causes a deadweight loss by reducing purchases of the taxed good. An income tax causes a dead- weight loss by reducing work efort and saving. A property tax causes a deadweight loss by reducing improvements to property. In each case, the size of the loss is determined by the sensitivity of the activity to its price. An income tax has a large deadweight loss if work efort increases signifcantly as wages rise or if saving is very sensitive to interest rates. In choosing taxes to fnance public services, communities should try to minimize deadweight loss. Taxes on land have a special appeal in this regard, because the supply of land is immobile and fxed by nature. As Henry George (1879) argued, a properly administered tax on land has no deadweight loss. To achieve this, the tax must not depend on the use of the land. If the tax is higher on improved land than on unimproved land, it will discourage productive use of land, causing a deadweight loss. Similarly, if the tax is higher on land used for commercial purposes than on land for homes and apartments, it will discourage com - mercial activity, causing a deadweight loss. George argued for a tax on the value of land in its highest and best use. In that case, the tax would not change if land use changed. A parcel tax could have the same benefcial efect as a land value tax (Oates and Schwab 2009). Te key is that the tax must not depend on the use of the land. Tat condition is easily satisfed by a tax on square footage. At frst look, the condition also seems to be satisfed by a fat tax per parcel. However, a fat tax tends to discourage the subdivision of large parcels, which in turn may discour- age residential development. To minimize deadweight loss, a parcel tax should be based on parcel size by levying a determined tax rate per square foot. To avoid deadweight loss, it should not be based on the use of the parcel or characteristics of any developments on the parcel. A parcel tax based on parcel size will have a much lower deadweight loss than an income, sales, or property tax. A uniform tax applied to all parcels in a community also reduces opportunities to export tax burdens to other communities. Te Desert Hot Springs parcel tax proposal is a good example of tax exporting. Te city proposed to increase the tax on vacant land, much of which belonged to residents of other communities, who could not vote on the proposal. Under a uniformity rule, the city could only increase its tax rate on vacant land if residents were also willing to increase the tax on their own parcels by the same amount. Equity Te equity principle is generally separated into two com- ponents: horizontal and vertical. A tax is horizontally equitable if two taxpayers in the same economic position pay the same tax. Te difcult issue is judging whether two taxpayers are in the same position. Income or wealth is commonly used to defne a taxpayer’s position. By that standard, a parcel tax is not horizontally equitable. Two taxpayers with the same income will pay diferent parcel In general, a tax will cause a deadweight loss whenever its application diminishes an economically productive activity. 17 Parcel Taxes as a Local Revenue Source in California www.ppic.org taxes if one owns a home and the other rents. Even if own- ership of land is used to determine a taxpayer’s economic position, a fat tax may not be horizontally equitable. Te main inequity concerns very small parcels created by subdivision. Tese parcels are too small to be developed by themselves and are ofen possessed by the owner of an adjacent parcel who treats the smaller parcel as part of the larger parcel. With a fat tax, the owner would pay a tax on both parcels. By contrast, with a tax on square footage, the total tax on the two parcels would be the same as if the parcels were one. Vertical equity concerns the tax burdens of taxpay- ers in diferent economic positions (Cordes 2011). Vertical equity is frequently judged according to the ability-to-pay principle. Higher-income taxpayers have greater ability to pay and should pay higher taxes, according to this prin- ciple. By that standard, a fat-rate parcel tax is not vertically equitable, because all homeowners pay the same regard- less of their income. A parcel tax based on square foot- age would be better, according to that standard, because wealthier homeowners tend to live on larger parcels. How- ever, the value of a homeowner’s parcel is an even better measure of ability to pay than the size of the parcel. Vertical equity can also be judged according to the beneft principle (Cordes 2011). By that standard, taxes should be related to the beneft that taxpayers receive from public services. A parcel tax looks relatively better accord - ing to this standard. If a community improves public services, it becomes more desirable for potential residents, increasing demand for housing there. If zoning regulations permit, this increased demand will increase the supply of housing. However, over the long run, that increase is limited by the supply of land. So enhancing a community’s public services tends to increase land values. Te increase occurs in both residential and nonresidential areas because nonresidential uses must compete for land with residential uses. Tus, landowners throughout the community are the primary benefciaries of improved public service. Accord- ing to the beneft principle, they should pay for those improvements, which they do if the improvements are fnanced by a parcel tax. Again, a tax on square footage is more equitable than a fat tax because the increase in value from better public services will be roughly proportional to parcel size. Te beneft principle is particularly appropriate for judging the parcel tax because of the way that it is levied. Te tax does not provide most of the funding for local public services. It fnances the marginal improvement of services over a baseline that is fnanced by other revenue. If the total beneft of an enhancement, measured by the willingness of each resident to pay for it, exceeds its cost, there is a strong case for fnancing it. Whether or not it is fnanced depends on voter support. Tat in turn depends on whether a voter’s willingness to pay exceeds the voter’s aversion to the additional tax. Te enhancement will receive voter support if, and only if, total benefts exceed total costs. Of course, perfect proportionality is impossible. But a good alignment of taxes and benefts will tend to make voter support for a proposal more likely when total benefts exceed total costs and less likely when benefts are less than costs. Equity is a difcult principle to apply, not only for the parcel tax but also for sales, income, and property taxes. Tax burdens under the income tax seem most consistent with the ability-to-pay standard. Sales tax and property tax burdens are also related to the ability to pay. By that standard, the property tax seems superior to the parcel tax, because it is not a fat rate per parcel or square foot but a tax based on the value of property. However, when judged according to the beneft principle, the parcel tax seems quite equitable relative to other taxes. A tax on square footage is more equitable than a fat tax because the increase in value from better public services will be roughly proportional to parcel size. Parcel Taxes as a Local Revenue Source in California 18 www.ppic.org Stability Taxpayers should be protected against sudden and unex- pected increases. During the 1970s, California housing prices increased rapidly, pushing up assessed values and property taxes. Tese increases explain at least some of the support for Proposition 13 (Fox 2003). In contrast, the par- cel tax is stable by its very nature, more so than the sales and income taxes. Simplicity Te parcel tax is also simple and less costly to administer than property, sales, and income taxes. In the case of the property tax, each county’s assessor maintains a roll of taxable properties with the names and addresses of parcel owners and various parcel characteristics, including land area. To distribute property tax revenue to local govern - ments, the county property tax administrator has a list of the local governments serving each parcel, which is used to generate property tax bills. Tose bills typically include additional charges, such as sewage and water. A parcel tax is just one more charge added to a parcel’s property tax bill. Transparency Regardless of whether the tax is fat or calculated per square foot, the basis for the parcel tax is objective and easy to understand. Unlike the income tax, there are no long forms to complete or complicated statutes to interpret. And, unlike a sales tax, every taxpayer knows his or her annual tax bill. A parcel tax bill comes once a year as an item listed on a parcel’s property tax bill. Unlike the sales tax, the parcel tax does not involve innumerable, small charges levied throughout the year. Integrity A tax that is easy to evade erodes public confdence in gov - ernment. Like the property tax, parcel taxes are difcult to evade. Parcels are easily identifed, parcel taxes are included on property tax bills, and governments may place a lien on properties with unpaid taxes. Te sales tax is also difcult to evade, but evasion is a problem with the income tax. Growth Tax revenue should grow in line with the cost of the services that the tax fnances. Unlike other taxes, a parcel tax’s base does not grow over time. Tus, it requires tax- rate adjustments to keep pace with rising costs. Some local governments have put in place automatic rate increases tied to the infation rate. Many school districts take a diferent approach. Tax proposals include a sunset provi- sion of usually four to ten years. When the tax expires, the district typically proposes another parcel tax that carries a rate that accounts for infation, enrollment growth, and changes in state aid. Tis approach is more fexible than automatic infation adjustments and also helps build a strong link between the district and its taxpayers. Shortcomings According to the seven principles of taxation, a uniform tax applied to square footage compares favorably with other taxes. However, it has one major shortcoming, and it concerns large parcels of little value. A parcel’s slope may be too steep for development, or the parcel may be far from an urban area and unft for agriculture. In such cases, the stream of future tax payments from a parcel tax may exceed any reasonable expectation of future income. Te application of the parcel tax in this case is akin to confs- cating land, which is unfair by any standard. If the tax were based on land value, this shortcoming could easily be overcome. However, a tax on value would violate the Proposition 13 limit on ad valorem taxes. Another approach would be to tax vacant land at a lower rate. But this approach would penalize development and violate the neutrality principle. Parcels of little value sufer According to the seven principles of taxation, a uniform tax applied to square footage compares favorably with other taxes. 19 Parcel Taxes as a Local Revenue Source in California www.ppic.org an inherent confict between neutrality and equity, which can only be resolved by a tax based on value. Taxing land fairly without taxing the value of land is challenging.Te difculty of this challenge depends on the tax rate. Every property has a tax capacity—that is, the tax rate that would threaten to reduce the property’s value to zero. A low tax rate is unlikely to exceed the capacity of any parcel. How - ever, as the tax rate rises, more properties become vulnerable. Te distribution of tax capacity will also vary by community. In suburban communities, tax capacities are likely to be relatively high across the board, allowing more leeway for tax rates than in more-rural areas, where large portions of land may have little value. Te distribution of tax capacity within communities is an empirical question worth investigating. Conclusion For the authors of Proposition 13, the parcel tax falls into that well-populated category of unintended consequences. Tey surely did not intend to limit property taxes only to create another way to tax real property. Proposition 13 capped property tax revenue, which arguably was excessive. But it also closed of the main avenue through which local governments could address the demands of their constituents. Te state has devised a formula for allocating property tax revenue among local governments. Yet no centralized formula can be expected to respond to the demands of constituents in such a large and diverse state as California. Some form of local, discre- tionary revenue is needed to supplement the property tax revenue allocated to each government. A local, discretion- ary source of revenue is also a crucial link between local governments and their taxpayers, an important factor in promoting an efective local public sector. Te parcel tax has assumed that role in several school districts, special districts, and cities. Tough currently lim - ited, parcel tax use is growing. Tat makes it increasingly important to ask whether it is an appropriate tax to play the role it is playing and what provisions might make it more acceptable. Tis report aims to advance that discussion. Because the parcel tax is a tax on land, it has many advantages over other taxes. Designed properly, it causes virtually no deadweight loss—that is, it does not diminish economic activity the way the sales and income taxes do. Moreover, it assigns taxes in line with benefts, a common standard of equity. Te essence of a properly designed parcel tax is not a fat tax but rather a tax per square foot of land, applied uniformly to all land uses. Such uniformity would also reduce opportunities for communities to export their tax burdens to other communities. Despite these advantages, a uniform tax on square footage has a major shortcoming. Some large parcels may have little value and thus be limited in their capacity to support a parcel tax. Te obvious solution is to base the tax on the value of land. However, a tax on value is prohibited by Proposition 13. Tat restriction is the primary obstacle the parcel tax must overcome. It is a challenge to tax land without taxing the value of the land. Tat challenge has been met in two ways. Te frst is to impose relatively low tax rates that do not exceed the tax capacity of low-value properties. Te Because the parcel tax is a tax on land, it has many advantages over other taxes. Designed properly, it causes virtually no deadweight loss—that is, it does not diminish economic activity the way the sales and income taxes do. The essence of a properly designed parcel tax is not a fat tax but rather a tax per square foot of land, applied uniformly to all land uses. Parcel Taxes as a Local Revenue Source in California 20 www.ppic.org second is to sacrifce uniformity—for example, by levy- ing a lower rate on vacant land. Te issue of uniformity is now before the state legislature, prompted by the Borikas decision, which overturned a parcel tax proposal that levied diferent residential and nonresidential rates. If the legislature insists on uniformity for school districts and extends that principle to cities and special districts, local governments’ use of the parcel tax will grow more slowly. A middle course would be to allow a few specifc deviations from uniformity, such as a lower tax rate for vacant land. Regardless of what the legislature decides, the state should acknowledge the signifcant role the parcel tax plays and begin to monitor it. In particular, the State Controller’s Ofce should collect data on parcel tax rates and revenue, and publish those data in its annual reports on cities, counties, school districts, and special districts. Notes 1 California Government Code, Section 53978. 2 California Government Code, Sections 50079 and 53717. 3 City and County of San Francisco v. Farrell, 32 Cal. 3d 47 (1982). 4 Taylor (2014) provides a fuller description of voter requirements for local taxes. 5 Taylor (2012) provides a fuller description of California’s prop - erty tax system. 6 Much of this material was covered in an earlier and more comprehensive study by Rueben and Cerdan (2003). Because this report focuses on broad-based taxes for general public services, the parcel taxes reviewed in this section are those levied by cities, school districts, and special districts, not the taxes levied by com - munity facility districts created by these other governments. 7 Coleman (2014) provides a detailed description of each revenue source. 8 Chavez and Freedberg (2013) and McGhee and Weston (2013) provide thorough summaries of parcel tax elections for California school districts. 9 California Government Code, Section 50079. 21 Parcel Taxes as a Local Revenue Source in California www.ppic.org References Baldassare, Mark, Dean Bonner, Sonja Petek, and Jui Shrestha. 2013. PPIC Statewide Survey: Californians and Teir Government. San Francisco: Public Policy Institute of California. Available at w w w.ppic.org/main/publication.asp?i=1044. California State Board of Equalization. 2014. A Report on Budgets, Workloads, and Assessment Appeals Activities in Cali- fornia Assessors’ Ofces 2012–2013. Sacramento. Available at www.boe.ca.gov/proptaxes/pdf/lta14003.pdf. California Tax Foundation. 2014. Piecing Together California’s Parcel Taxes: An In-Depth Survey of Local Taxes on Property. Sacramento. Available at w w w.caltaxfoundation.org/reports /parceltax-summary.pdf. Chavez, Lisa, and Louis Freedberg. 2013. Raising Revenues Locally: Parcel Taxes in California School Districts 1983–2012. Oakland, CA: EdSource. Available at edsource.org/wp-content /publications/pub13-ParcelTaxesFinal.pdf. Coleman, Michael. n.d. California Local Government Finance Almanac. Available at w w w.californiacity fnance.com/index .php#VOTES. Coleman, Michael. 2014. Te California Municipal Revenue Sources Handbook. Sacramento: League of California Cities. Cordes, Joseph J. 2011. “Vertical Equity.” In Te Encyclopedia of Taxation and Tax Policy , 2nd ed., edited by Joseph J. Cordes, Robert D. Ebel, and Jane Gravelle. Washington, DC: Urban Institute Press. Coupal, Jon. 2013. “Te Shameful History of Parcel Taxes.” Howard Jarvis Taxpayers Association, February 10. Available at w w w.hjta.org/california-commentary/shameful-history-parcel - t a x e s /. Doerr, David R. 2000. California’s Tax Machine: A History of Taxing and Spending in the Golden State. Sacramento: California Taxpayers Association. Fox, Joel. 2003. Te Legend of Proposition 13: Te Great Califor- nia Tax Revolt. N.p.: Xlibris Corporation. George, Henry. 1879. Progress and Poverty: An Inquiry into the Causes of Industrial Depression and of Increase of Want with Increase in Wealth. New York: Doubleday, Page & Co. Hill, Elizabeth G. 2000. Reconsidering AB 8: Exploring Alterna- tive Ways to Allocate Property Taxes. Sacramento: Legislative Analyst’s Ofce. Available at www.lao.ca.gov/2000/020300 _ab8/020300_ab8.pdf. Lang, Bree, and Jon Sonstelie. 2014. “Te Parcel Tax as a Local Revenue Source for California Public Schools.” Unpublished manuscript, Xavier University and University of California, Santa Barbara. Available at econ.ucsb.edu/~jon/documents /LangSonstelieParcelTax _4_18_14.pdf. Manatt, April, Kimia Mizany, Peter Detwiler, and Elvia Diaz. 2010. What’s So Special about Special Districts? A Citizen’s Guide to Special Districts in California. 4th ed. Sacramento: California Senate Local Government Committee. Available at www.csda .net/wp-content/uploads/2013/04/WhatsSoSpecial_2010.pdf. McGhee, Eric, and Margaret Weston. 2013. Parcel Taxes for Education in California. San Francisco: Public Policy Institute of California. Available at www.ppic.org/content/pubs/atissue /AI_913EMAI.pdf. Oates, Wallace E., and Robert M. Schwab. 2009. “Te Simple Analytics of Land Value Taxation.” In Land Value Taxation: Teory, Evidence, and Practice , edited by Richard F. Dye and Richard W. England. Cambridge, MA: Lincoln Institute of Land Policy. Ramseth, Luke. 2014. “Desert Hot Springs: With Parcel Tax Measure, City Hopes to Avoid Second Bankruptcy.” Riverside Press Enterprise , May 2. Rueben, Kim S., and Pedro Cerdan. 2003. Fiscal Efects of Voter Approval Requirements on Local Governments. San Francisco: Public Policy Institute of California. Available at www.ppic.org /content/pubs/report/R _103KRR.pdf. Sonstelie, Jon. 2014. “California’s Parcel Tax.” Working paper, Lincoln Institute of Land Policy. Available at www.lincolninst .edu/pubs/dl/2460_1808_ SonstelieWP14JS1.pdf. Taylor, Mac. 2012. Understanding California’s Property Taxes. Sacramento: Legislative Analyst’s Ofce. Available at www.lao .ca.gov/reports/2012/ta x/property-ta x-primer-112912.pdf. Taylor, Mac. 2014. A Look at Voter-Approval Requirements for Local Taxes. Sacramento: Legislative Analyst’s Ofce. Available at w w w.lao.ca.gov/reports/2014/fnance/local-taxes /voter-approval-032014.pdf. Parcel Taxes as a Local Revenue Source in California 22 www.ppic.org 22 About the Author Jon Sonstelie is an adjunct fellow and Bren fellow at the Public Policy Institute of California and a professor of economics at the University of California, Santa Barbara. His areas of expertise are public fnance and urban economics, with a focus on school fnance. He coauthored the California School Finance Revenue Manual , which explains the complexities of public school funding in California, and he developed the PPIC School Finance Model for simulating the efects of various school funding formulas. His work includes studies of school budgets and student achievement, state fnance and the Master Plan, school resources and academic standards, and the efect of public school quality on private school enrollment. He was previously a research fellow at Resources for the Future. He holds a PhD in economics from Northwestern University and a BA from Washington State University. Acknowledgments I am grateful to Michael Coleman and EdSource for sharing their data on parcel tax elections. For comments on an earlier draf of this report, I am also grateful to Chas Alamo, Michael Coleman, Rob Gutierrez, Patrick Murphy, Marianne O’Malley, Terri Sexton, Lynette Ubois, Paul Warren, and Joan Youngman. Te Lincoln Institute of Land Policy provided important support in the early stage of this research. Any errors in this work are my own. 23 Parcel Taxes as a Local Revenue Source in California www.ppic.org Board of Directors DONNA LUCAS, CHAIRChief Executive Ofcer Lucas Public Afairs MARK BALDASSAREPresident and CEO Public Policy Institute of California RUBEN BARRALESPresident and CEO GROW Elect MARÍA BLANCOExecutive Director, Undocumented Student Legal Services Center University of California Ofce of the President BRIGITTE BRENAttorney LOUISE HENRY BRYSONChair Emerita, Board of Trustees J. Paul Getty Trust WALTER B. HEWLETTMember, Board of Directors The William and Flora Hewlett Foundation PHIL ISENBERGVice Chair Delta Stewardship Council MAS MASUMOTOAuthor and Farmer STEVEN A. MERKSAMERSenior Partner Nielsen, Merksamer, Parrinello, Gross & Leoni, LLP KIM POLESEChairman ClearStreet, Inc. THOMAS C. SUT TONRetired Chairman and CEO Pacifc Life Insurance Company PPIC is a public charity. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public ofce. PPIC was established in 1994 with an endowment from William R. Hewlett. Copyright © 2015 Public Policy Institute of California. All rights reserved. San Francisco, CA Mark Baldassare is President and Chief Executive Ofcer of PPIC. Donna Lucas is Chair of the Board of Directors. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source. Research publications refect the views of the authors and do not necessarily refect the views of the staf, ofcers, or Board of Directors of the Public Policy Institute of California. Library of Congress Cataloging-in-Publication Data are available for this publication. I S B N 978 -1-5 8213 -16 0 -3 PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 Telephone 415.291.4400 Fa x 415 . 2 91. 4 4 01 PPIC SACRAMENTO CENTER Senator Ofce Building 1121 L Street, Suite 801 Sacramento, California 95814 Telephone 916.440.1120 Fax 916.440.1121 Additional resources related to fscal/governance reform are available at www.ppic.org. The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research." ["post_date_gmt"]=> string(19) "2017-05-20 09:42:29" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_415jsr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:42:29" ["post_modified_gmt"]=> string(19) "2017-05-20 09:42:29" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_415JSR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }