Reconsidering the Optional Single Sales Factor

Reconsidering the Optional Single Sales Factor M AC TAy lo r • l e g i s l A T i v e A n A l y s T • M Ay 26, 2 010 an LaO RepORt 2 LegisLative anaLyst’s Office an LaO RepORt ExEcutivE Summary The February 2009 state budget agreement changed the apportionment formula used to determine California taxable income for firms that also operate in other states. While the cur- rent formula considers the location of firms’ sales, property, and payroll, starting in 2011 firms will have the option to consider only their sales. This policy is intended to encourage firms to produce in California and sell into other states. In this report, we examine the rationales for different approaches to apportionment and evidence from California and other states on how changes to apportionment laws affect both economic activity and tax revenue. Our findings indicate that: ➢ A formula with a higher weight on sales and lower weights on property and payroll promotes job growth to some extent. ➢ With most states’ formulas now based only on sales, the old formula that used property and payroll could put some California producers at a competitive disadvantage. ➢ Allowing firms to choose their formula every year arbitrarily favors some firms over others. We recommend that the state require all firms to use the single sales factor, which would help the state’s competitiveness while limiting the cost to the budget. LegisLative anaLyst’s Office 3 an LaO RepORt 4 LegisLative anaLyst’s Office an LaO RepORt introduction The February 2009 state budget agreement In this report, we examine the rationales changed the apportionment formula used to deter- for different approaches to apportionment and mine California taxable income for firms that also evidence from California and other states on operate in other states. While the current formula how changes to apportionment laws affect both considers the location of firms’ sales, property, economic activity and tax revenue. We then and payroll, starting in 2011 firms will have the compare the state’s new law to other states’ laws option to consider only their sales (“single sales”). and recommend some changes. FEaturES oF StatE apportionmEnt FormulaS Firms report profits only at the national State Laws. Despite the name, UDITPA has never level as opposed to the state level, so states been a federal law. Its policy rationale was that have devised a process known as “apportion- income should be apportioned based on the fac- ment” to determine what fraction of a multistate tors of production (property representing capital firm’s profits they can tax. The term implies that and payroll representing labor) and a sales factor the states in which a firm operates divide its to represent market presence. (Sales are counted profits up so that the sum of the taxable profits in the state where a product is consumed, not claimed by each state is equal to the firm’s total. produced.) Under UDITPA, each factor had a However, the reality is that each state in which weight of one-third. California adopted UDITPA the firm operates performs its own calculation in 1966. using its own method. In all 45 states that levy a In 1993, California modified its formula by corporate income tax, taxable profits at the state adopting a “double-weighted” sales factor. That level are based on the percentages of the firm’s is, the weights on both property and payroll are national property, payroll, and/or sales located in 25 percent, but the weight on sales is that state. States set their own relative weights on 50 percent. Many other states similarly reduced property, payroll, and sales. A firm that operates the weights on payroll and property and in- in states that use different formulas may find that creased the weight on sales in the 1980s and the sum of its taxable profits in the states it oper- 1990s. The rationale for this policy was to induce ates is higher or lower than its overall national mobile firms that can produce in one state but profits. In response to differing state formulas, sell into others (typically manufacturers) to locate firms have an incentive to use tax planning to facilities and employees in the state. minimize their overall state tax bill. Alternatives to Apportionment. The only The three-factor apportionment framework feasible alternative to an apportionment method comes from the Uniform Division of Income is called separate accounting, which is used for Tax Purposes Act (UDITPA) that most states by the federal government. At the state level, it adopted following a 1957 meeting of the Na- would mean that a firm would have to establish tional Conference of Commissioners on Uniform a California affiliate for tax purposes, and its LegisLative anaLyst’s Office 5 an LaO RepORt state taxable profits would be defined as receipts sales shipped from California into states where earned by its California affiliate (including “sales” the shipping firm has no nexus are “thrown to its own operations in other states) minus de- back” and counted as California sales for corpo- ductible expenses incurred in California. No state rate tax purposes. Because the throwback rule uses this method, as it would require much more pertains to sales, it is more significant the higher bookkeeping than the current system and would the state’s sales factor is relative to the other two encourage firms to game the system by putting factors. If a state has a throwback rule, firms that unrealistic “prices” on their internal transac- ship from that state will not have the incentive tions so that their receipts are disproportionately described above to avoid establishing nexus in booked in jurisdictions with the lowest tax rates. states that they ship into. For example, under Nexus. While there are no federal laws current law if a court rules that a firm that ships governing most aspects of state apportionment, from California into Utah has no nexus in Utah, federal case law prevents states from levying cor- then Utah will not be able to impose its corpo- porate income taxes on firms with no “nexus” (ba- rate tax on the firm. Instead, California will be sically, physical presence) in the state. As a result, able to count the firm’s Utah sales as California firms with minimal facilities in a state but signifi- sales for tax purposes. As an alternative example, cant sales may have an incentive to avoid any tax if the firm initially had nexus in Utah, it could in liability by reducing their presence in the state to some cases lower its overall tax bill by closing the point where they no longer have nexus. its Utah facilities and eliminating this nexus. This Throwback Rules. California is one of several would likely be the case if Utah had a higher tax states with a throwback rule, which means that rate than California. How apportionmEnt aFFEctS rEvEnuES and incEntivES: a SimplE ExamplE Apportionment’s impact on a firm’s tax bill The firm has taxable profits of $8 million in depends on the distribution of its sales, property, Alabama and $4 million in Georgia for a total and payroll among the states it operates in. For example, a hypothetical firm with $10 million of Figure 1 total profits operates in just two states: Alabama Hypothetical Two-State Firm: and Georgia. Its property, sales, and payroll are Factor Allocation split as shown in Figure 1. (Dollars in Millions) Alabama’s apportionment formula puts equal Alabama Georgia weights of one-third on each of the three fac- Share Share tors. In contrast, Georgia’s formula uses only the Amount of Total Amount of Total sales factor and puts weights of zero on property Payroll $45 90% $5 10% and payroll. Figure 2 shows how the two states Property 360 90 40 10 Sales 60 60 40 40 would compute the firm’s taxable profits. 6 LegisLative anaLyst’s Office an LaO RepORt of $12 million, yet its Figure 2 national profit is just Hypothetical Two-State Firm: $10 million. This arises Taxable Profits Exceed Total because the firm’s facili- (Dollars in Millions) ties are disproportionate- Alabama Georgia ly in Alabama where the 1/ Payroll, 1/ Property, 1/ Sales 100% Sales 3 3 3 formula uses property Share Weight Subtotals Share Weight Subtotals and payroll, and its sales Payroll 90% 33% 30% 10% — — are disproportionately in Property 90 33 30 10 — — Georgia where the for- Sales 60 33 20 40 100% 40% mula uses only sales. Apportionment ratio 80% 40% Switching the factor Share of $10 million in profits $8 $4 percentages from Fig- Taxable Profits—Both States Total $12 ure 1 so that 90 percent of the firm’s payroll and property and 60 percent Figure 3 of its sales are in Georgia Hypothetical Two-State Firm: instead of Alabama leads Taxable Profits Less Than Total to a very different result. (Dollars in Millions) As shown in Figure 3, Alabama Georgia the firm’s taxable profits 1/ Payroll, 1/ Property, 1/ Sales 100% Sales 3 3 3 at the state level add up Share Weight Subtotals Share Weight Subtotals to less than its national Payroll 10% 33% 3.3% 90% — — profits. Property 10 33 3.3 90 — — When states use dif- Sales 40 33 13.3 60 100% 60% ferent formulas (all else Apportionment ratio 20% 60% equal), firms have an Share of taxable profits $2 $6 incentive to locate their Taxable Profits—Both States Total $8 facilities in states that put more weight on sales and sell into states that put more weight on property lower if its facilities are concentrated in Georgia, and payroll. In this example, the firm’s tax bill is which creates a policy problem for Alabama. LegisLative anaLyst’s Office 7 an LaO RepORt tHEory and EvidEncE on apportionmEnt Theory Is Inconclusive, Evidence Links jobs by about 3.3 percent. Another used data Profits to Sales. There is no consensus among from 1978 to 1999 and accounted for some ad- economists or other policy experts as to the ditional factors such as sales and property taxes, theoretically appropriate factors or weights on and reached a similar conclusion. A third, using each factor. As such, the assignment of equal data from 1987 to 1996 and accounting for some weights under UDITPA was somewhat arbitrary. additional factors such as special tax incentives There is, however, some empirical evidence on and public spending, found that a higher sales the statistical relationship of profits to the three factor increases the amount of spending on busi- factors. A recent study of 11,000 European firms ness facilities and equipment. In addition to these for which 2004 data on profits, property, pay- studies, a number of other studies using data for roll, and sales were all available concluded that either just one state or just one year have pro- the apportionment formula that best fit the data duced mostly similar results. would use weights of 27.5 percent on property, Results of a 2005 simulation using the 7 percent on payroll, and 65.5 percent on sales. California-specific Dynamic Revenue Analysis A similar calculation for a smaller sample of U.S. Model (DRAM) suggest that a mandatory single firms suggests weights of roughly 45 percent on sales factor would create jobs on net—roughly property, 5 percent on payroll, and 50 percent one job for each $17,400 (in 2001 dollars) of ini- on sales. tial revenue loss. (The simulation also accounted Evidence Links Higher Sales Factor With Job for the impact of state spending cuts that would Growth. In general, the evidence suggests that be needed to offset the revenue loss.) This figure increasing the weight of the sales factor produces suggests that mandatory single sales could pro- a small but noticeable increase in economic duce an eventual net gain of about 40,000 jobs activity. Several academic studies which have based on the Franchise Tax Board’s (FTB’s) latest covered all states over a decade or more and cost estimates. The DRAM estimate of the state controlled for other factors that affect economic and local revenue feedback effect from increased activity besides corporate taxes have found that sales, property, and income taxes stemming from a higher sales factor (all else equal) is associated the job gains was about 14 percent of the initial with more economic development. For example, revenue loss, which is consistent with the general one used data for all states from 1978 to 1994 empirical evidence on these effects. These esti- and found that on average, switching from a mates should be interpreted with caution as they double-weighted sales factor to a single sales are based on the 2001 state economy and incor- factor would increase a state’s manufacturing porate a lot of assumptions, but we believe that they are reasonable ballpark figures. 8 LegisLative anaLyst’s Office an LaO RepORt StatE’S Formula will cHangE in 2011 Optional Single Sales Factor. As part of the each of the firm’s property and payroll but just February 2009 budget agreement, starting with 20 percent of its sales. the 2011 tax year, firms will be able to choose The figure shows that a California-based firm between the current double-weighted sales for- would be able to reduce its tax bill by 60 percent mula and a new formula that uses the single sales by switching from the current double-weighted factor, ignoring the property and payroll factors. sales factor formula to the new single sales factor. (The new law does not apply to banks and agri- In contrast, Figure 5 (see next page) shows cultural or mining firms. These firms will contin- the same calculation for an out-of-state firm that ue to use the UDITPA equal-weighted formula as has relatively high sales in California (14 percent) they do under current law.) Firms will be able to compared to its shares of property and payroll decide each year which of the two formulas they (4 percent each). This firm would save $44,200 want to use for that year. under the current formula with double-weighted Other Apportionment Changes. The state sales, so it will elect not to use the new formula. also modified some other provisions of the ap- Optional Formulas Will Favor Some Firms portionment law to partially offset the cost of Without a Clear Rationale. Firms will benefit the formula change. It clarified the definitions of from being able to switch from one formula to nexus and gross receipts, broadened the defini- the other depending on whether they are hav- tion of sales attributable to “unitary” groups of ing a good year or a bad year. Consider what affiliated businesses, and changed the treatment would happen if a firm records a $10 million loss of sales of services. instead of a $10 million profit. Now it wants to maximize the taxable loss, which can be de- Impact of 2009 Policy Change ducted against its taxable income for prior years Below, we discuss the likely effects of last year’s or future years. Figure 6 (see next page) shows policy change to give businesses a choice between this calculation for the two firms from the previ- the two formulas. Some Firms Will Figure 4 Benefit From New Single Sales Versus Double-Weighted Sales: Formula. Consider the California-Based Firm choice between double- Double- Single weighted sales and single Weighted Sales sales for a hypothetical Payroll 80% 80% firm with $10 million Property 80 80 Sales 20 20 of national profits that sells nationwide but has California apportionment ratio 50% 20% most of its operations in × Total U.S. profits (millions) $10.0 $10.0 = California taxable profits (millions) 5.0 2.0 California (see Figure 4). California Tax Payment at 8.84 Percent $442,000 $176,800 The state has 80 percent LegisLative anaLyst’s Office 9 an LaO RepORt ous examples and a third firm that has balanced to the current optional formula will reduce state factors: 16 percent each of its property, payroll, revenues by $900 million annually by 2012-13 and sales in California. when it is completely phased in. (This includes The California-based firm will now switch to the definitional changes, which would be expect- double-weighted sales to maximize its loss, and ed to raise revenue on their own.) The estimated the out-of-state firm will likewise switch to single revenue loss would be somewhat lower— sales. In contrast, the “balanced” firm records $50 million to $100 million annually—if all firms the same taxable loss under either formula and were required to use the single sales factor in- thus does not benefit at all from being able to stead of retaining the option to stick with double- choose. Similarly, a firm that operates only in weighted sales. The FTB used its database of California cannot benefit from choosing its for- corporate returns from 2006 to construct these mula because it does not use the apportionment Figure 5 method in the first place. Single Sales Versus Double-Weighted Sales: In other words, the op- Out-of-State Firm tional single sales factor Double- Single gives a bigger benefit to Weighted Sales out-of-state firms than to Payroll 4% 4% balanced multistate firms Property 4 4 Sales 14 14 or California-only firms. Revenue Impact. The California apportionment ratio 9% 14% FTB estimates that the × Total U.S. profits (millions) $10.0 $10.0 = California taxable profits (millions) 0.9 1.4 switch from mandatory California Tax Payment at 8.84 Percent $79,560 $123,760 double-weighted sales Figure 6 Single Sales Versus Double-Weighted Sales: Money-Losing Firms California-Based Firm Out-of-State Firm “Balanced” Firm Double- Double- Double- Weighted Single Weighted Single Weighted Single Payroll 80% 80% 4% 4% 16% 16% Property 80 80 4 4 16 16 Sales 20 20 14 14 16 16 California (CA) 50% 20% 9% 14% 16% 16% apportionment ratio × Total U.S. loss (millions) $10.0 $10.0 $10.0 $10.0 $10.0 $10.0 = CA loss (millions) $5.0 $2.0 $0.9 $1.4 $1.6 $1.6 California tax savings $442,000 $176,800 $79,560 $123,760 $141,440 $141,440 Difference $265,200 $44,200 $0 10 LegisLative anaLyst’s Office an LaO RepORt estimates, and its assumptions about profit growth an option to go with a double‑weighted until 2012‑13 are consistent with the forecasts sales factor starting in 2011. Firms will be used to develop the state’s 2010‑11 budget. required to use the double‑weighted for‑ mula for at least three years before switch‑ Other States’ Use of Optional Formulas ing back to the equal‑weight formula. State tax systems that allow firms to choose their apportionment methods are not common, ➢ South Carolina allows firms to petition the and little evidence is available on the impact of state’s revenue agency to use an alternate allowing firms to choose. Four states currently formula if the firm believes that the state’s have optional formulas of some type in place. prescribed formula (double‑weighted sales through 2010, single sales for 2011 ➢ Missouri is the only state currently that and beyond) does not fairly represent the allows an annual election between single firm’s income. It appears that the depart‑ sales and the traditional three‑factor ment typically approves these requests. formula. The state has not conducted a The state has conducted no studies on the study on the impact of this policy that effects of allowing this option. was adopted in 1973. Colorado had a longstanding policy of allow‑ ➢ Utah changed its law in 2005 to allow ing firms to choose annually between double‑ firms to choose between the traditional weighted sales and equal weights, but replaced three‑factor formula and a double‑ this system with mandatory single sales in 2008. weighted sales factor for five years at a Alternatives to Optional Formulas. A few time. The law has not been in effect long states use different mandatory formulas for differ‑ enough for them to draw conclusions ent sectors, usually by mandating a higher sales about its impact. factor for sectors that typically sell into other ➢ New Mexico currently requires the states. An example of this is Maryland, which traditional equal‑weighted three‑factor uses a single sales factor for manufacturers but formula but is going to give manufacturers double‑weighted sales for all other firms. Recommendations California has been criticized at times for forces the state to either raise other taxes or cut having high costs of doing business. The single spending on public services. In view of this, we sales factor would reduce those costs for mobile discuss below the merits of mandatory versus firms who sell into national or world markets optional formulas, the single sales factor versus and are more of a flight risk than firms who the double‑weighted sales factor, and the timing sell only into the California market. The results of the implementation of any changes. of the DRAM simulations and other empirical Choice of Apportionment Formula. Allow‑ evidence suggest that a higher sales factor (all ing a choice between single sales and double‑ else equal) generates some employment growth. weighted sales arbitrarily favors firms with dispro‑ The tradeoff is that it reduces revenue and portionately high or low California sales relative LegisLative anaLyst’s Office 11 an LaO RepORt to property and payroll. These firms will benefit taxable profits in excess of its actual profits. For most from switching their formulas around from this and the reasons noted above, we recom- year to year depending on whether they report a mend that the state use a single sales factor. We net profit or a loss, and will thus pay a lower tax also recommend that the state keep the throw- rate over the business cycle than multistate firms back rule to avoid the economic distortions with more evenly distributed factors or firms that discussed earlier. operate only in California. Given these concerns, Timing. Given the state’s ongoing budget we recommend that firms be required to stick to shortfall, we have recommended that the Leg- a single formula. islature consider delaying the implementation Single Sales or Double-Weighted Sales? The of a number of tax policy changes that reduce strongest case for single sales concerns confor- state revenues enacted as part of the last two mity to other states’ policies. If all states impose years’ budgets. The state currently faces a nearly a corporate income tax with a single sales factor $20 billion budget shortfall for 2010-11. The fol- and a throwback rule, then a multistate firm’s lowing budget in 2011-12 will be challenging as total taxable income at the state level will be well given the expiration of billions of dollars in equal to its total nationwide income. As such, temporary taxes. Consequently, we recommend the playing field between multistate firms and that the Legislature delay any changes in appor- California-only firms will be level. The same tionment policies for two years. would also be true if all states used the double- Overall Approach weighted formula or the equal-weight three-fac- In 2009, the Legislature signaled its intent tor formula. However, the dominant formula now to follow other states in switching to a single among the large states is single sales: Texas, New sales factor. While there is a good case for a York, Virginia, Georgia, Massachusetts, Illinois, mandatory single sales approach, providing it as Michigan, and Ohio all use single sales while an option creates clear disparities among busi- only Florida, New Jersey, and North Carolina still nesses. We recommend that the state replace use the traditional or double-weighted formulas. the optional single sales factor with a mandatory Pennsylvania is in between, with a 75 percent single sales factor beginning in 2013. This would weight on its sales factor. increase state General Fund revenues Conformity with other states would prevent by about $215 million in 2010-11 and about California firms from being placed at a competi- $700 million in 2011-12 and 2012-13 (due to the tive disadvantage. Under mandatory double- delay in implementation). Thereafter, our recom- weighted sales, a California producer that sells mendation would increase revenues by up to into states with single sales could well have total $100 million each year. LAO Publications This report was prepared by Justin Garosi, and reviewed by Michael Cohen. The Legislative Analyst’s Office (LAO) is a nonpartisan office which provides fiscal and policy information and advice to the Legislature. To request publications call (916) 445-4656. This report and others, as well as an E-mail subscription service, are available on the LAO’s Internet site at www.lao.ca.gov. The LAO is located at 925 L Street, Suite 1000, Sacramento, CA 95814. 12 LegisLative anaLyst’s Office