When Intel named Chandler, Ariz., as the site of its next U.S. fabricating plant in July, politicians touted the announcement as validation that changing the state's corporate income tax formula had been a wise move. And Intel made it easy for politicos to take credit for the new "fab." Tom Franz, Intel's vice president and general manager of fab/sort manufacturing, was quoted as saying, "Passage of the sales factor was clearly one of the key reasons we picked the state of Arizona," as a home for the plant.

Intel's $3 billion dollar fab now being built in Chandler is exactly the type of manufacturing facility that states are competing for when they change their corporate income tax to hit sales, rather than payroll or property. Arizona is among several states changing tax formulae to boost economic development. But do changes in tax rules bring states the business upswing they seek?

Intel made it clear that tax incentives were not the only reason the company chose Chandler. Businesses like Intel analyze a host of factors determining where to locate facilities. In addition the sales tax factor, Franz cited "science and technology" in discussing Intel's decision calculus. "In addition to a favorable tax climate," Franz said, "Arizona has strong research universities and a commitment to developing a deeper infrastructure in science and technology, including improvements in all levels of education."

This underscores the difficulty of sorting out cause and effect in examining which factors drive firm relocation decisions. In fact, research suggests states may give up more than they gain when they tinker with tax formularies to encourage corporate investment.

Formula for attraction

Traditionally, states have levied corporate taxes using an apportionment formula in which sales, payroll and property were weighted equally, explains Sanjay Gupta, professor of accountancy at the W. P. Carey School of Business. During the 1990s, this method for apportioning state corporate income tax liability began changing.

Today, only 11 states equally weight the three apportionment factors -- sales, property and payroll -- says Gupta. In 1982, 34 states used the equally weighted apportionment formula.

Michael Mazerov, a senior fellow at the Center for Budget and Policy Priorities, notes that 15 states now use a single-sales-factor apportionment formula for state corporate income tax, which means the states tax sales only. All other states with a corporate income tax use some form of super-weighted sales formula where, for instance, sales account for 50 percent of a corporation's tax liability while payroll and property account for just 25 percent each.

That is the formula Arizona had been using but, in May, the state adopted legislation that will, by 2009, allow corporations to opt for a new tax modus operandi that would cut tax liability for companies with facilities in the state. Under the new law, corporations may use the old apportionment formula or choose instead to have sales count for 80 percent of the company's tax liability while payroll and property make up the remaining 20 percent.

"The popular speculation is that if you change your tax formula, more businesses will locate in your state," Gupta says. Companies whose business is global are likely to log the majority of their sales outside the state, out of reach of its tax code. Property and payroll, which are rooted in place, would not be taxed under the new formula. Although this lowers the corporate tax, lawmakers hope the state eventually will enjoy revenue growth fueled by the rise in employment that results from the business lured to the state by the tax break. At least, that's the argument used by proponents of such tax formula changes. Gupta's research, however, indicates the benefits are questionable. "What we are beginning to find is that revenues go down," he says.

The price of progress

Gupta has recently teamed with Jeffrey Gramlich from the University of Southern Maine and Mary Ann Hofmann of Andrews University in Michigan to produce a soon-to-be published paper entitled "Empirical Evidence on the Revenue Effects of State Corporate Income Tax Accounting Policies." In it, the authors note that recent evidence contradicts those who argue that states will eventually gain from changes in apportionment formulae.

"For example, during the three years following its move to a single sales factor apportionment formula, Illinois suffered a $200 million loss in corporate income tax revenues, as well as a decline in manufacturing jobs," the team writes. "Ohio, Pennsylvania and Massachusetts have had similar experiences."

Theoretically, the revenue impact of changes in the corporate tax blueprint "could go either way" from a tax collections perspective, says Kelly Edmiston, a senior economist with the Federal Reserve Bank of Kansas City. "In practice, it tends to be negative," he adds, explaining that there are reasons for revenue decline.

For one thing, Edmiston says, companies with no people or facilities in a state are protected by a federal law that maintains simply having sales in a state is not sufficient to create taxable nexus (or connection). "If the company just has sales in the state, you can't tax their income. They have to have some kind of payroll and property in the state," he adds.

This means while companies with payroll and property in the state see their tax liabilities go down, sales-only companies may not be seeing their taxes go up. In fact, some aren't paying any taxes at all. In that case, "There are no offsetting gains" for the state, Edmiston explains.

In addition, Edmiston's research shows that corporate payroll and property tend to increase in a state and sales tend to decrease. "That makes sense. That's just companies responding to tax incentives," he says. But, in the states making tax-structure changes to court companies, "payroll and property are not going to be taxed, so even though they're going up, there is no tax gain on them."

Gupta adds, "Changing the apportionment formula by placing heavier weights on the sales factor is aimed at benefiting in-state firms relative to out-of-state firms." He agrees with Edmiston that tax revenue could rise or fall, depending, in part, on the proportion of revenues derived from in-state versus out-of-state firms. "To the extent a state derives most of its corporate income taxes from in-state firms, it stands to lose revenues right off the bat," he says. "If, on the other hand, a state derives most of its revenues from out-of-state firms, it may see a revenue increase in the short run. But on a dynamic basis, all bets are off because firms are likely to shift their presence strategically so as to manage their tax burdens."

Other gains can offset corporate tax break

Still, Edmiston is quick to point out that the decline in corporate income tax isn't the whole picture. "There are other tax bases affected," he says, maintaining that when payroll taxes go up, a state might expect personal income taxes to go up, also.

This played into the reasoning Edmiston used when asked to project probable revenue outcomes for the state of Georgia, which recently made a switch to the single-factor formula. In 2003, Edmiston anticipated that the move would be revenue negative if the change was implemented in 2004 but that the state would subsequently gain $45 to $47 million in additional revenue by 2006, driven, in part, by a rise in personal taxes.

Edmiston and other researchers have identified a gain in manufacturing employment when states shift away from uniformly factored apportionment. However, based on his research for the Center for Budget and Policy Priorities, Michael Mazerov has found "no compelling evidence that the switch to single sales factor is a powerful incentive" for businesses to locate facilities in one state versus another.

Massachusetts adopted the single-factor formula taxing sales only in 1995. Between then and 2004, Mazerov tracked manufacturing job growth and retention in the five states that used this tax formula.

According to Mazerov, although virtually every state in the U.S. lost manufacturing jobs during that time, three single-factor states -– Iowa, Texas and Nebraska -– did better than average in job retention. One state matched the national median, and one state, Massachusetts, did very badly. "Massachusetts suffered the fifth steepest decline in manufacturing jobs of all 50 states," Mazerov says, adding that his findings might be "weak evidence" for proponents of single-sales-factor taxation.

But if you look at a longer time period and a larger sample of states, the evidence points toward a different conclusion. Mazerov did another study in which he tracked manufacturing job growth and retention for the eight states that had single-sales-factor apportionment in place from December 2001 to December 2004. "In the more recent period, five of the eight single-sales-factor states have done worse than the average state in manufacturing job retention," he says.

Which states were the manufacturing-job winners between 1995 and 2004? According to Mazerov's findings, "The three states that had the best manufacturing-job growth over that period still use the equally weighted three-factor formula."

Over the '95 to '04 time period, Mazerov also looked at how single-factor states fared in attracting large plants. Three of the five states with single-factor tax formulas failed to attract a single large plant, which Mazerov defines as one costing $700 million or more.

And what about Intel? Does it really place plants where tax codes reward in-state payroll and property investments?

Mazerov looked at all Intel's major plant location decisions between 1990 and 2004. He found that even though two states, Iowa and Missouri, had single-factor formulas in place for the entire time, Intel had put eight-and-half times more investment in non-single-factor states.

 

According to the Arizona Republic, the state's "Department of Revenue calculated that the change would cost state coffers $90 million to $100 million when fully implemented in 2010."  On balance, however, that figure doesn't include business development gains that may be spurred on by the change in tax policy.

Even if the policy change does inspire new business investment, Edmiston points out that corporate income taxes were originally designed to be "benefit taxes." That is, the state levies taxes to pay for services that benefit the corporation, such as roads to the factory and education for factory workers. "You want to tax the business in proportion to the benefit it receives from government services," he says. "That's efficient from an economist's point of view."

By lowering tax liability for property and payroll, the state moves away from taxing companies relative to the benefits they receive, Edmiston says. "It's no longer a benefit tax system," he maintains.

So why do policymakers adopt these policies? Edmiston understands the state lawmakers' point of view. States do this to "level the playing field" with other states competing for corporate investments, he says.

Competition may be the name of the game in business, but Gupta points out that when governments compete with each other, it's not the same as business competition.

"When businesses compete with each other you get efficiencies that come out of the whole system," he says. When states compete for business with tax formulas, you can get less money coming in to pay for the government services corporations use.

"In this type of competition, everybody loses," Gupta says.