Skip to main content

State Enactments of the "Single Sales Factor" Tax Incentive Have Had Little Impact on Intel Corp.’s Major Plant Location Decisions

I never made an investment decision based on the tax code. . . If you are giving money away I will take it. If you want to give me inducements for something I am going to do anyway, I will take it. But good business people do not do things because of inducements, they do it because they can see that they are going to be able to earn [at least] the cost of capital out of their own intelligence and organization of resources.

Former Alcoa CEO Paul O’Neill at his confirmation hearing to be President George W. Bush’s first Secretary of the Treasury, January 17, 2001

Background and Summary

Intel Corporation, the world’s largest semiconductor manufacturer, is actively lobbying the Arizona legislature to enact a law implementing a corporate tax incentive that allegedly makes a state an attractive location in which to engage in manufacturing. The incentive is known as the “single sales factor formula.” The Arizona House approved its version of the single sales factor legislation, H.B. 2139, on March 7th. The Senate version, S.B. 1043, could be brought to the floor for a vote at any time.

The single sales factor formula is a departure from the traditional method by which a state determines the share of the nationwide profit of a corporation upon which it will levy its corporate income tax. The single sales factor formula determines that share solely with reference to the share of the corporation’s nationwide sales located in the state — with the location of sales deemed to be the state in which goods are delivered to purchasers. If Arizona were to adopt the single sales factor formula, and if Intel could arrange to deliver all of its computer chips to its customers outside of Arizona, then Intel would pay little if any corporate income tax to the state. [1] Under present Arizona law, in contrast, Intel’s substantial physical presence in Arizona — it has two major chip fabrication plants in the state — ensures that Intel makes significant income tax payments to the state in years in which it is profitable.

Intel is hinting that if Arizona wishes to be seriously considered as a location for the company’s future computer chip fabrication plants, it must enact the single sales factor (hereafter, SSF) legislation. [2] Intel is also currently lobbying Oregon local officials for a 15-year renewal of a special program that reduces property taxes on Intel’s Oregon facilities. Intel is telling public officials there that this legislation is needed to ensure that Oregon is in the running for future Intel plants.

Well-aware that they are in competition with other jurisdictions, Arizona officials are understandably concerned that their failure to enact the single sales factor legislation sought by Intel could harm their ability to attract future Intel facilities and the high-paying jobs that come along with them. Nonetheless, an examination of Intel’s past plant location decisions provides powerful evidence that the presence or absence of a single sales factor formula in a state’s corporate tax code has not had a significant impact on where Intel has chosen to site its major facilities or on how many jobs these facilities generated:

  • In the last fifteen years, Intel made $1.9 billion in major plant investments in single sales factor states after they had enacted SSF legislation. (See the table on page 8.) During this same period, Intel made major plant investments of at least $16.45 billion in states that did not have single sales factor in effect at the time — even though there were states with SSF in which the plants could have been sited. In short, Intel has placed more than eight and one-half times as much investment in non-SSF states as in SSF states since 1990. [3]
  • Intel’s Oregon facilities are the largest of any state’s, representing nearly $11 billion in cumulative investment.[4] Of that amount, only $400 million was announced after August 2001, when Oregon enacted the first of two bills that are phasing-in the single sales factor formula.
  • Intel’s Oregon employment today is actually slightly below what it was the year before the phase-in of SSF was enacted. Intel’s Oregon employment peaked in 2000 at approximately 16,000 workers, and today it stands at 15,500. [5] In other words, Oregon has yet to show any net Intel job gains from its reduction of Intel’s Oregon income taxes via its enactment of the single sales factor formula. [6]

A more detailed recounting in the body of this report of the history of Intel’s plant siting decisions provides substantial additional evidence that neither the single sales factor formula itself nor state corporate tax policy in general has been a significant driver of Intel’s location choices. Rather than being motivated to invest in a state by the presence in the tax code of a single sales factor formula, it appears that the presence of a major Intel investment in a state — and the political influence such investment brings — motivates Intel to seek single sales factor legislation to reduce its tax liability.

Arizona vs. Oregon (vs. California )

Individuals outside the company do not have access to sufficient information about Intel’s customers and operations to determine whether Arizona’s adoption of a single sales factor formula would entirely eliminate Intel’s corporate income tax liability to the state. There is little doubt, however, that SSF would cut Intel’s Arizona tax payments substantially. Intel is hinting that if Arizona wishes to be seriously considered as a location for the company’s future computer chip fabrication plants, it must enact the SSF legislation. In a recent speech to the Arizona Technology Council, Intel CEO Craig Barrett warned: “The corporate tax issue in Arizona [i.e., the absence of SSF] is in fact absolutely designed to demotivate us to make further investments in the state.”[7]

In lobbying aggressively in Arizona for SSF legislation, Intel is hoping to replicate its earlier success in Oregon, where the company’s largest U.S. facility is located. Intel was a prime mover behind that state’s enactment of 2001 and 2003 laws that began phasing-in the single sales factor formula in 2003 and will fully implement it in 2008.

With SSF in hand in Oregon, Intel has turned its attention to obtaining expanded property tax breaks for its Oregon facilities. Even as it is suggesting to Arizona policymakers that they must enact SSF if the state wishes to attract future chip fabrication plants, Intel has suggested to Oregon policymakers that they must renew the state’s “Strategic Investment Program” (SIP) if Oregon wishes to be in contention for future Intel investments. Under SIP, the company is eligible for abatement of all property taxes on new investments in chip-making equipment in excess of $100 million. Since Intel’s newest chip plants cost in excess of $2 billion — with machinery composing a large share of the total investment — the tax savings from SIP are substantial.[8] According to its Oregon spokesperson, the company is lobbying for a 15-year renewal of SIP to “set the stage for making Oregon able to compete for Intel’s long-term investments.”[9] This is, of course, the same justification Intel is offering for its demand that Arizona enact SSF.

Intel has indicated that it is likely to announce the location of its next major chip fabrication plant in the second half of 2005. If it is placed in the United States, the company already has a strong incentive to site it in Oregon or Arizona, because Intel already owns an adequate amount of vacant land on the same “campuses” where its existing plants are located. If both Oregon and Arizona accede to the tax reductions sought by Intel, one of them could see the plant go to the other state. Arizona could lose substantial corporate tax receipts in making the single sales factor formula available to all corporations already present in the state yet not capture the new Intel plant that appears to be a significant motivation for the SSF legislation.

Finally, it is worth noting that Oregon and Arizona are not the only two states that Intel has been pitting against each other for additional tax breaks. Intel also has been lobbying for the enactment of the single sales factor formula in its headquarters state of California for several years now. Intel’s chief tax lobbyist testified in favor of SSF to the California legislature as recently as February 9, 2005.

Has Single Sales Factor Really Influenced Intel’s Location Decisions?

Unlike with most tax incentives, the tax savings from SSF are not limited to profits associated with new investments; the formula works in such a way that it automatically reduces corporate tax payments flowing from existing investments as well. [10] The enactment of SSF provides a substantial windfall tax savings to any corporation like Intel that serves a national (or in Intel’s case, worldwide) market from a relative handful of production locations — whether or not such a company ever makes a new investment in the SSF state.

As its push for SSF legislation in Arizona, Oregon, and California indicates, Intel is more than willing to use it political clout in states in which it is a major employer to lobby for changes that will provide it with such a valuable tax windfall. Nonetheless, there is substantial evidence that the presence or absence of the single sales factor formula in a state has had very little influence on where Intel has chosen to locate its major facilities up to now:

  • Intel sited its first chip fabrication plant outside of its home state of California in Oregon, in 1976. (Intel calls these plants “fabs” and refers to them by number, such as Fab 12 and Fab 22 in Arizona.) At that time, Iowa and Missouri both had the single sales factor formula in effect. A document on the history of Intel’s Oregon operations available on the company’s Web site suggests that non-tax considerations drove Intel to choose Oregon:

Oregon had landed on Intel’s radar when a member of the Tektronix board, interested in expanding Oregon’s high technology universe, called Intel’s co-founder, Gordon Moore. “You should consider Portland. It’s got a good quality of life and a good labor pool,” Keith Thomson, a former Intel Oregon Site Manager, recalls. “We were looking for a place within an hour or two by airplane from San Jose . . .” [11]

  • Intel opened its first Arizona fab in 1980 — again rejecting the single sales factor states of Iowa and Missouri.
  • By the time Intel broke ground on its second Arizona fab, in 1994, there were four states that had enacted a single sales factor apportionment formula — Iowa and Missouri had been joined by Nebraska and Texas. Once again, Intel’s investment had not been attracted to any of the available SSF states.
  • In November, 1996, Intel announced plans to build a $1.3 billion fab in Texas, a single sales factor state. Over 500 acres of land were purchased near Ft.Worth, and construction was actually begun. In January, 2000, however, construction was suspended, and the company announced permanent cancellation of Fab 16 in February, 2003. This appears to be the only U.S. fabrication plant that Intel has ever cancelled after beginning construction. Subsequent to suspending construction of Fab 16 in January, 2000, Intel invested $1.5 billion in a Colorado fab, $2 billion in a New Mexico fab, $2 billion in an Oregon fab, and $2 billion in the current (nearly completed) upgrade of Arizona Fab 12. None of these states had the single sales factor formula in effect when these investment decisions were made. In other words, Intel cancelled a fab in a single sales factor state and subsequently placed major fabs in non-SSF states.
  • Intel has invested approximately $3 billion in its two New Mexico fabs, Fab 11 and Fab 11X. New Mexico is not a single sales factor state. By the time Intel announced plans to build Fab 11X (May 2000), the company had six single sales factor states to choose from: Iowa, Missouri, Nebraska, Texas, Illinois, and Massachusetts.
  • Massachusetts is the only single sales factor state in which Intel has sited a major fabrication plant after the state had adopted the single sales factor formula. Even here, however, it is questionable that the availability of SSF played a major role in Intel’s location decision. Intel bought an existing fab from Digital Equipment Corporation in 1998, which it then upgraded to a $1.5 billion investment. It seems likely that its decision to locate in Massachusetts was driven less by the state’s adoption of SSF and more by Intel’s ability to acquire a facility at a good price from a financially-troubled corporation — which it could then put into operation more quickly than a “greenfield” plant built from the ground up. This hypothesis is supported by the fact that when a similar opportunity arose in a non-SSF state, Intel took advantage of it: in early 2000, Intel bought a Colorado fab from Rockwell International and upgraded it with a $1.5 billion investment.
  • By the time (2000) Intel announced plans to build its newest fab — D1D at its Ronler Acres, Oregon campus — there were seven states with SSF in effect: Iowa, Missouri, Nebraska, Texas, Massachusetts, Illinois, and Connecticut. Oregon had not yet enacted SSF legislation, but Intel chose Oregon over these other seven states anyway.
  • As noted above, Intel’s Oregon facilities are the largest of any state’s, representing nearly $11 billion in cumulative investment. [12] Of that amount, only $400 million (a conversion to other uses of its mothballed Fab 15, in Aloha), was announced after August 2001, when Oregon enacted the first of two bills that are phasing-in the single sales factor formula.
  • Intel’s Oregon employment today is actually slightly below what it was the year before the phase-in of SSF was enacted. Intel’s Oregon employment peaked in 2000 at approximately 16,000 workers, and today it stands at 15,500. [13] In other words, Oregon has yet to show any net Intel job gains from its reduction of Intel’s Oregon income taxes.
  • Intel has been lobbying actively for the enactment of the single sales factor formula in Arizona at least as far back as 2001. Despite the Arizona legislature’s failure to enact the legislation, Intel went ahead with a $2 billion upgrade to its Arizona Fab 12, which it announced in February 2003.
  • In the last fifteen years, the only major plant investments Intel made in single sales factor states after they had enacted SSF legislation were the previously-noted $1.5 billion Fab 17 in Massachusetts and the $0.4 billion conversion of Oregon Fab 15. (See the table on page 8.) During this same period, Intel made major plant investments of at least $16.45 billion in states that did not have SSF in effect at the time — despite the fact that as early as 1992 there were four states with SSF in effect and additional states enacted SSF throughout the period. In other words, Intel has placed more than eight and one-half times as much investment in non-SSF states as in SSF states since 1990.

It is not surprising that the presence or absence of SSF does not appear to have had a significant impact on Intel’s major plant location decisions. The state corporate income tax is an almost trivial expense for a large corporation like Intel. In the late 1990s, before Oregon enacted the first of two bills to phase-in SSF, Intel was paying an average of $51 million in corporate income taxes annually to that state. In 2001, its Oregon payroll expenses alone were $1.65 billion.[14] If a company is going to spend more than 30 times as much for labor as it is going to pay in state corporate income taxes, then the availability and price of the labor is likely to have far more influence on its location decisions than is the tax liability.

Two other issues provide further evidence that state corporate income tax policies have not been a significant factor in Intel’s plant location decisions:

  • Intel has kept its corporate headquarters in California and has placed major facilities in Oregon, Arizona, and Colorado. All four of these states mandate that corporations calculate their income taxes using “combined reporting.” Combined reporting is a form of state corporate income taxation that thwarts many types of corporate tax-avoidance strategies. [15] All other things being equal, a corporation with options as to where to locate its major plants would tend to avoid combined reporting states. The fact that Intel has made major investments in all four is further evidence that state corporate tax policy is not a major factor in the company’s location decisions.
  • Conversely, given Intel’s propensity for southwestern locations (the company has facilities in California, Utah, Arizona, New Mexico, and Colorado), one might expect to find significant Intel facilities in Nevada, a state with no corporate income tax whatsoever. But there do not appear to be any significant Intel offices or plants in Nevada — still further evidence that state corporate income taxes are not significant location factors for Intel.

Conclusion

None of the evidence presented in this report concerning Intel’s past plant location decisions proves that Arizona would not be disadvantaging itself in attracting future Intel facilities if it fails to enact the single sales factor legislation the company is seeking. It does suggest, however, that any such disadvantage would arise from factors other than Arizona’s having created an objective, economically-significant disincentive for Intel to invest within Arizona’s borders.

Intel is the 53rd largest corporation in the Fortune 500. It earned more than $10 billion in profits between 2001 and 2003. It paid on the order of 2 percent to 4 percent of those profits in income taxes to all the states in which it was taxable in those years. [16] In seeking to reduce its state corporate tax liabilities further by pushing Arizona, Oregon, and California to enact single sales factor legislation, Intel is not doing so because it must in order to remain profitable and competitive. It is doing so to marginally increase those profits.

Who Will Pay to Educate the High-Tech Entrepreneurs of Tomorrow?

A full-page ad in the May 2, 1997 New York Times taken out by City College of New York featured one of its most famous alumni, the legendary former CEO of Intel Corporation, Andrew Grove. The ad read, in part:

In 1957, I arrived in New York as a refugee from Hungary. Three days later, barely able to speak English, I took the subway up to City College. Within hours, I was registered — with a scholarship.

I am forever grateful to CCNY — they gave me a future. And today, the school continues to offer a high-quality education to deserving students, many of them first-generation immigrants. [Emphasis in original]

While Grove quite rightly can be grateful to CCNY for the education he received there, some gratitude is perhaps also due to the taxpayers of New York City whose tax dollars paid for that education (which was free at the time). Grove owes some thanks to the taxpayers of California as well; as recounted in his 2001 autobiography, Swimming Across, Grove “went through graduate school on scholarships” at the University of California at Berkeley — another public institution.

During the recent state fiscal crisis, higher education took some of the deepest cuts in state financial support of any category of state services. One result was that tuition and fees at state universities grew 10 percent annually in each of the last three academic years, according to the College Board. This was double the rate of increase in the preceding three years. Public university costs continue to escalate; for example, on March 10, 2005, the Arizona Board of Regents approved tuition increases of as much as 9.8 percent for Arizona residents attending state universities next year. This is occurring in a state that ranked 40th in the share of its 2004 high school class that enrolled in post-secondary education last Fall and that increased its 2003-2004 state university tuition by nearly 40 percent. As one state’s higher education commissioner told the Wall Street Journal, “A one-year double-digit rise is one thing, but when you talk about multiple years, states need to ask ‘Are we able to sustain access to all the people who need an education?’ ”

Intel Corporation has been aggressive in seeking corporate income tax cuts, property tax abatements, and public subsidies (such as industrial revenue bonds) in states like Arizona in which it has a major economic and physical presence. It may reasonably be asked whether such tax-reduction efforts — and similar efforts by major multistate corporations throughout the United States — have reached a point where they are now impairing the ability of those governments to fund the education systems that will “give a future” to our nation’s up-and-coming Andrew Grove-s.

Whether reducing Intel’s state tax burden further is in either the company’s or Arizona’s long term interest is open to question. Senior Intel managers have been quite outspoken recently in warning that the inadequacy of the public education system in the United States is a threat to the country’s economic competitiveness. Intel CEO Craig Barrett recently wrote:

By any objective measurement, the U.S. K-12 educational system is broken and must be fixed if our children and grandchildren are going to enjoy the standard of living we have come to expect. The unfortunate truth is that the average U.S. high school graduate ranks in the bottom 10 percent of the industrialized world in math and science competition. The “No Child Left Behind Act” is a good start. Measurement and accountability are essential to improvement. We should also implement the recommendations of the Glenn Commission to improve math and science education by recruiting good teachers, giving them the right tools and training, and paying them what they’re worth, based on achievement of results. [17]

Worrisome questions can be raised about whether Arizona is adequately preparing its young people to fill the high-technology jobs of the future in companies like Intel. It is estimated that Arizona ranked 40th among the states in the percentage of the high school class of 2004 that enrolled in post-secondary education last Fall.[18] It ranked 38th in the share of its 2004 high school graduating class that took a college-level Advanced Placement exam while enrolled in high school and 32nd in the share that actually passed at least one such exam.[19] It ranked 42nd in the share of its 8th-grade students that scored “proficient” and above on the National Assessment of Educational Progress 2003 Reading Assessment.[20] Adequate financial resources will be needed for Arizona to make progress on these indicators. It appears that the leadership of Intel perceives little connection between these very real problems and its drive to reduce the company’s state and local tax burdens as much as possible. Arizona policymakers hopefully will weigh these tradeoffs carefully before taking final action on single sales factor legislation.

Table 1:
Major Intel Investments in Single Sales Factor and Non-SSF States, 1990-2004
(classification of SSF and Non-SSF at the time investment announced)

Amt Invested

Years

Location and description

SSF States

 

 

$1.5 billion

98-00

Hudson, MA campus; acquisition and upgrade of Fab 17

$0.4 billion

03

Upgrade of former Fab 15, Aloha, OR

$1.9 billion

 

Total, Single Sales Factor States

Non-SSF States

 

 

$1.0 billion

93

Construction of Fab 11, Rio Rancho, NM

$0.7 billion

94

Conversion of Fab D1A to Fab 15, Aloha, OR

$1.3 billion

94-96

Construction of Fab 12, Chandler, AZ

$2.2 billion

94-96

Acquisition of Ronler Acres Campus, OR, and constr. of Fab D1B/20

$1.5 billion

98-99

Construction of Fab D1C, Ronler Acres Campus, OR

$2.0 billion

00-01

Construction of Fab 22, Chandler, AZ

$1.5 billion

00-01

Acquisition and construction of Fab 23, Colorado Springs, CO

$2.0 billion

00-02

Construction of Fab 11X, Rio Rancho, NM

$2.0 billion

00-03

Construction of Fab D1D, Ronler Acres Campus, OR

$0.25 billion

00-01

Construction of RP1 Lab, Ronler Acres Campus, OR

$2.0 billion

03-05

Upgrade of Fab 12, Chandler, AZ

$16.45 billion

 

Total, Non Single Sales Factor States

End Notes

[1] The mechanics of the single sales factor formula, its impact on state corporate tax collections, the reasons it is alleged to be an economic development incentive, and the empirical evidence that it has not, in fact, had a significant impact on job creation in states that have adopted it are discussed at length in a previous Center report. See: Michael Mazerov, The “Single Sales Factor Formula” for State Corporate Taxes: A Boon to Economic Development or a Costly Giveaway? , Center on Budget and Policy Priorities, revised September 2001. For purposes of this report, it is sufficient to know that the single sales factor formula is alleged to be a potent incentive for manufacturers to locate in states that have adopted it.

[2] Arizona reporters are certainly interpreting the company’s position in that way. See: Ed Taylor, “Intel Hints at EastValley Expansion, East Valley Tribune, February 18, 2005; Associated Press, “Intel Considers Chandler Expansion,” Arizona Daily Sun, January 16, 2005.

[3] 1990 was selected as the starting point for this comparison because there is little information available on Intel’s Web site concerning the value of plant investments made before that time. All of Intel’s major plant investments prior to 1990 were in California and Oregon, which were non-SSF states at the time. (California remains a non-SSF state today.)

[4] Aliza Earnshaw, “Intel Seeks Extension of Tax Breaks,” The Business Journal of Portland, February 23, 2005.

[5] Intel’s current Oregon employment is stated to be approximately 15,500 at www.intel.com/community/oregon/index.htm. Its peak employment of 16,000 in 2000 is listed in a December 2004 slide presentation by Steven R. Grant, Vice President, Technology & Manufacturing Group, available at www.oregonbusinessplan.org/pdf/Steve%20Grant%20Presentation.pdf.

[6] Oregon is slightly more than half way toward phasing in SSF. An apportionment formula that gives an 80 percent weight to sales went into effect in May 2003. (Sales had previously been weighted 50 percent.) Sales will be weighted 90 percent effective in July 2006, and the state will move to a 100 percent, or “single” sales factor formula in July 2008.

[7] Quoted in Associated Press, “Intel Considers Chandler Expansion,” Arizona Daily Sun, January 16, 2005.

[8] Under SIP, Intel makes (smaller) payments in lieu of property taxes to the local governments abating its property tax, and it continues to pay property tax on land and buildings.

[9] Quoted in Aliza Earnshaw, “Intel Seeks Extension of Tax Breaks,” The Business Journal of Portland, February 23, 2005.

[10] See the source cited in Note 1.

[11] Most of the information in this section of the report was obtained from various documents available from three sections of Intel’s Web site: “Oregon in Your Community” (home page at www.intel.com/community/); a “Virtual Press Kit on Intel’s Worldwide Manufacturing Operations” (available at www.intel.com/pressroom/kits/manufacturing\index.htm); and Intel’s press release archive (available at appzone.intel.com/pressroom/PressReleases.asp). Because of the large number of documents reviewed to prepare this section of the report, most are not individually cited here; copies are available from the author.

[12] Aliza Earnshaw, “Intel Seeks Extension of Tax Breaks,” The Business Journal of Portland, February 23, 2005.

[13] See Note 5.

[14] ECONorthwest, Economic Impacts of Intel’s Oregon Operations, February 2003. Available on Intel’s Web site at www.intel.com/community/oregon/downloads/ECONorthwest.pdf. Oregon corporate income tax payments on p. 43, and Oregon payroll payments on p. 18.

[15] See: Michael Mazerov, Closing Three Common Corporate Income Tax Loopholes Could Raise Additional Revenue for Many States, Center on Budget and Policy Priorities, Revised May 21, 2003.

[16] Robert S. McIntyre and T.D. Coo Nguyen, State Corporate Income Taxes 2001-2003, Citizens for Tax Justice, January 2005.

[17] Craig R. Barrett, “America Must Choose to Compete,” Semiconductor Industry Association Quarterly Newsletter, Winter 2005.

[18] College Board, Advanced Placement Report to the Nation 2005, January 2005, p. 7.

[19] Advanced Placement Report to the Nation 2005 , Appendix C.

[20] Jennifer Sloan McCombs, Sheila Nataraj Kirby, Heather Barney, Hilary Darilek, and Scarlett J. Magee, Achieving State and National Literacy Goals, a Long Uphill Road, Rand Corporation, 2004, p. 47.