Wednesday, September 21, 2016

Fred Thompson: The LRO’S Research on Measure 97

Fred Thompson checks in with the forth in a series of posts on Measure 97 (IP28).

Proponents of Measure 97 deny that most of its burden will be shifted forward to consumers and that it will thereby reduce the progressivity of Oregon’s tax system. Moreover, they have consistently questioned the expertise and impartiality of the Legislative Revenue Office (LRO) for reporting that both claims are likely correct. In my opinion, their criticism of the LRO is entirely unjustified. If anything, the LRO has bent over backwards to treat M-97 in an even-handed fashion.

It is very hard to address these issues without getting wonky, but I’ll start by saying that economists who study tax questions agree that taxes on sales (gross receipts, turnover, value-added, retail-sales, etc.) are for the most part shifted forward to consumers. We would also insist that this is not necessarily the case with all of the taxes that businesses remit to the government. For examples, the incidence of taxes on property, business income (profit or earnings), payroll, or the extraction of raw materials, tends to fall on business owners or is shifted back to factor suppliers.

Nevertheless, M-97 is different from any known sales or gross receipts tax. Typical sales taxes discriminate along jurisdictional lines or by type of product. M-97 proposes to discriminate on the basis of the seller’s identity, levying the tax ONLY on the sales of certain C-corporations, those with Oregon sales in excess of $25 million. The LRO tried to deal with this issue by making the portion of the tax that would be shifted forward to consumers proportional to the concentration of affected businesses in each class of business, so that, if big C-corps represented 50 percent of sales in that market segment, only 25 percent of the tax would be shifted forward, 90 percent of sales = 80 percent shifted, and 100 percent = all. Because affected firms tend to dominate the business categories in which they operate, the LRO concluded that overall more than 60 percent of the tax take of ≥$3 billion would be shifted forward from business owners to consumers.

Maybe this is an overestimate, maybe even a big one. Perhaps, most of the burden of M-97 really will be born by the shareholders of extremely large, mostly out of state companies or even out of state consumers.  We really cannot know for sure until M-97 is put into effect – which is in my opinion a bug not a feature. But on theoretical grounds, it can be argued with some justification that the LRO has, if anything, underestimated the amount of forward shifting M-97 will induce. In imperfectly competitive markets, economists usually assume that market leaders are price setters: that, when their cost of goods-sold increases, they will raise their prices and, depending on their market power, the increase may be less than the cost increase, but may also be MORE. Since market power is usually assumed to be a function of concentration, this is essentially the logic that underlies the LRO’s very reasonable position.

What the LRO’s position rules out is the possibility of over-shifting and, especially, over-shifting due to the behavior of the competitive fringe, in this instance the businesses not facing the M-97 induced cost increase. The standard assumption in competitive analysis is not, as proposed by M-97 supporters, that they will undercut the price of the dominant businesses in the area, but will, instead, tend to meet the price increase of the market leaders. Where this is the case, the price increase borne by consumers as a result of a tax may and, in some cases, will exceed the tax take.

Moreover, this is generally what the evidence shows happens. The closest analogy to M-97, that I know of, occurs at borders, where one jurisdiction has a higher embedded excise than another, typically on things like motor fuels, alcohol, or tobacco products. Not surprisingly, prices just across the border in the higher taxing jurisdiction tend to be higher on average than in low-tax jurisdiction, but not by as much as in the rest of that jurisdiction, where various studies show that between 80 and 400 percent of the tax take tends to be recouped by businesses. What might seem surprising, however, is that prices just across the border in the low-tax jurisdiction are also typically higher than elsewhere in that jurisdiction – and, of course, retail booze, cigarettes, and gas sales are probably characterized by more price taking than one is likely to see in a lot of affected business segments under M-97.

My point is not that the LRO has underestimated the amount of tax shifting likely to occur under M-97. Rather, they have made a prudent, middle-of-the-road estimate, which is their job. The extreme estimates? They come from M-97’s supporters, and are based neither on economic theory nor independent evidence.

Monday, August 29, 2016

Measure 97: Any Pinocchios Yet?

 Fred Thompson checks in with an third post on Measure 97 (IP28). 

So far, my criticisms have been directed to the pro-97 camp – to claims like Gov. Brown’s that “there is a basic unfairness in our tax system that makes working families pay an increasing share for state and local services.” Not so. Oregon has long had one of the US’s least regressive state and local tax systems and recent changes have increased its progressivity. For example, the state already has the nation’s highest capital-gains tax rate. And, even if it were so, Measure 97 wouldn’t make the tax system fairer, but would cause ‘working families’ to pay more for government services and to pay for them more regressively.

I acknowledge that most of the claims I have criticized are at least somewhat arguable – mostly wrong, not Pinocchios. For example, the pro-camp claims that some of the $3 billion or so that Measure 97 will collect each year “will come out of CEO pay, some out of dividends, some out of consumers in other states.” But, they acknowledge, “Some will come from some of us.” That’s basically the position of the Legislative Revenue Office, except that they estimate that the “some [that] will come from some of us” is about 2/3 the total, that most of the rest will be shifted to the IRS, with the remainder borne by shareholders (about $200 million a year, which is enough to explain a lot of opposition) and corporate employees (about $100 million).

Think what it would look like, were Measure 97 truly a tax on corporations: total Oregon profits run about $20 billion a year. The share earned by C-corps, the only businesses subject to the tax under Measure 97, is about half that, or $10 billion, of which 70 percent comes from corporations with more than $25 million a year in Oregon sales, or about $7 billion, from whom the state currently collects about $375 million a year. If Measure 97 were to increase the tax take from these corporations by $3 billion a year, we would be talking about an average effective state corporate income tax (CIT) rate of nearly 50 percent – five times the highest state CIT rate in the US. Because gross receipts taxes are recorded as ordinary expenses and reduce ‘‘earnings from operations,’’ remittances to the state would cut federal CIT obligations by about $1 billion (given the 35 percent federal CIT rate), but, even so, this implies an average combined CIT rate of over 60 percent.  Bizarrely, however, this tax only hits businesses with low profit rates (as a proportion of turnover); really high profit businesses would continue to be unaffected by it. How credible is that? And, if it’s credible, imagine what this could do to the state’s economy.

This is also an area where Measure 97’s opponents have crossed the line. Many of them condemn Measure 97 as unfair because unprofitable businesses, which appear to have no ability to pay, are nonetheless obliged to pay the tax. They express similar concerns with respect to low-margin, high-turnover businesses, for example, grocery stores. However, as Fox, Luna, and Murray observe, economists generally do not share these concerns, “because the base is not intended to be profits and the tax is likely shifted forward to consumers.” Arguably, these anti-claims merely take the pro-side’s argument about who pays seriously, but it’s probably not how this would play out.

Of course, gross receipts taxes, like most taxes, create tax wedges. In this instance the wedges affect both interstate and intrastate commerce. They arise because some of the goods and services providers that are subject to the tax must compete with those that are not. This may induce them to shift activities out of state or to purchase out-of-state inputs to avoid the tax. Compared with a retail-sales tax say, Measure 97’s distortions are potentially more severe, both because of pyramiding and because of the discrepancy between the tax rate faced by C-corps (2.5 percent) and the implicit rate (zero) faced by pass-through entities. But this looks to be a matter of degree, not of kind.

Measure 97’s supporters make a big deal of the fact that only about a 1,000 businesses will be directly affected by it: “It will raise taxes on only the largest corporations, affecting one-quarter of 1 percent of the companies doing business in Oregon.” (Of course, many more, probably most, will be indirectly affected by it through their supply chains – for example, most businesses in the state will face higher utility costs). However, $3 billion doesn’t grow on trees – to give some idea of the magnitude of what we are talking about here, that’s about what a 6.5 percent retail sales tax would produce. If the businesses affected by Measure 97 weren’t major players in the state’s economy, a 2.5 percent tax wouldn’t yield $3 billion. They are, The 1,000 corporations directly affected by Measure 97 employ 40 percent of the state’s business labor force, account for over half of the state’s value added and over half of its private-sector wages. In other words, they comprehend the state’s most productive enterprises and those that pay the best salaries.

Following up on the last point, in performing its analysis of the effects of Measure 97, both the Legislative Reference Office and the analysts at PSU presumed that the financial services industry would be largely unaffected by it. That is the case for two reasons. First, this industry is heavily weighted to pass-through entities, which are exempt from the gross receipts tax under Measure 97 (see Figure 4 from Owen Zidar).

 Second, both explicitly assumed that Oregon would exclude the sale of real property, investment receipts, including interest, dividends, and capital gains, and sales of financial instruments, including bonds, mortgages, debentures, etc., in calculating gross receipts. Consequently, most financial service companies would only be on the hook for their management fees. And, given that the gross receipts tax under Measure 97 is an alternative minimum tax and that, where financial transactions are excluded, financial service companies tend to be high-margin, low-turnover businesses, this means that, for the most part, they will be unaffected by Measure 97 – C-corps in the financial services industry tend to pay fairly high corporate income taxes now, they would continue to do so under Measure 97.

Excluding sales of financial instruments is standard practice where financial services are concerned, since financial transactions vastly exceed state product. This issue hadn’t come up in Oregon previously because it’s irrelevant to defining the corporate-income tax base or to determining the existing alternative minimum tax, which is capped at $100 thousand. But it is inconceivable that the legislature wouldn’t address this issue if Measure 97 were to pass. Otherwise the results could be pretty scary. Interestingly, there is one financial service industry where this exclusion typically doesn’t apply, insurance. Presumably, Measure 97’s gross receipts tax would apply to insurance premiums purchased by Oregonians – indeed, the LRO estimates that 20 biggest insurers doing business in Oregon would see their state tax remittances increase from average of $200,000 to $3.5 million per annum.

This quasi-anomaly is partly due to the distinction between apportionment on the basis of “market” or “activity.” Oregon’s CIT apportionment is currently governed by the taxpayer’s principal business activity (PBA) code on their federal tax returns – and these codes are largely based on the NAICS code, which classifies business activities as goods or services – goods are currently apportioned on a market basis (defined in terms of their customers location) and services on an activity basis (where the bulk of the work is performed). Consequently, PBA codes control which apportionment method applies to a taxpayer when computing Oregon’s business tax. Most finance services, for example, take a service code. However, this isn’t necessarily the case. Grant Thornton offers the following illustration: “A company that processes financial transactions using advanced technology has all of its employees located in Oregon. It determines for Oregon tax purposes that a ‘technology’ NAICS code is most appropriate. Businesses that use the ‘technology’ code apportion receipts to Oregon based on the location of the business’s customers, which often results in apportionment of less than 100 percent.” In contrast, if the company chooses a ‘financial’ code, the apportionment percentage would be based on the company’s payroll in Oregon, which for this business would be 100 percent. However, because the federal PBA code disclosure doesn’t affect the computation of federal tax, the feds don’t require businesses to choose a code with any precision or consistency (one important caveat aside, where the selection of NAICS codes affects tariffs paid on goods/services crossing international borders, customs authorities strictly constrain gaming by multinational corporations of NAICS codes) – this is important because state administration of income taxes piggybacks on federal reporting, and playing with these codes is one of the ways that C-corps currently shift their tax liabilities away from higher-tax jurisdictions, like Oregon, to those with lower CIT rates, like Washington.

Gov. Brown has proposed to fix this anomaly, should Measure 97 pass, by basing all business taxes on the location of the customer buying the service, rather than the location of the company selling the service.  Unfortunately, the state doesn’t have a mechanism for tracking in-state sales or following them back up the value chain. Right now, we pretty much rely on businesses to tell us what their Oregon sales are. Unfortunately, it is easy for businesses with out-of-state sales to fiddle this figure. Indeed, our inability to track in-state sales is one reason for Measure 67’s adoption of dollar limits for its alternative minimum tax. Consequently, the claim made by the pro-97 camp, that, because the measure targets sales rather than profits, businesses will find it harder to avoid Measure 97 taxes, may, in fact deserve a couple of Pinocchios.

Monday, August 1, 2016

More Background on IP28 (Measure 97?), cont.

 Fred Thompson checks in with another post on IP28. 

Is it true, as asserted by OCPP and Our Oregon, “that businesses benefited greatly from seismic changes to Oregon’s property tax system in the 1990s,” which caused a “shift in property taxes away from businesses and onto households.” This claim is an important part of the justification for IP-28 (Measure 97?), at least insofar as enactment of this measure is supposed to redress corporate tax gains made at the expense of the citizenry at large, but, in fact, it is probably not true that businesses benefited disproportionately from Measures 5 and 50.

So, let’s start with what is certain.

It is certain that Oregon used to be a high property tax state and that now, because of Measures 5 and 50, it is a middling one. Prior to Measures 5 and 50, Oregon property taxes averaged about 5 percent of disposable income; nowadays they’re more like 3.5 percent, as depicted in the following figure from the Research Section of the Oregon Department of Revenue. As a result of these measures, Oregon’s property tax growth is also a lot more stable (less volatile).

It is also certain that in 2014-15 taxes on residential properties accounted for the lion’s share of property-tax payments, about 57.6 percent, and for about the same share of the state’s net assessed property value. However, it is also certain that Measure 50 caused assessed value to deviate from real market value (or, perhaps, more correctly restored the discrepancies between assessed value and real market value that were commonplace before Measure 5) – over time the faster the growth in real market value, the greater the discrepancy. Evidently, the real market value of residential property has grown faster than the values of other kinds of property. As a result, the average effective tax rate (tax payments/real market value) on residential property is only now 1.06 percent and on owner occupied housing it’s even lower, while the effective tax rate on all other property is significantly higher at 1.32 percent.[1] If anything, this suggests that Oregon’s current property tax system is biased in favor of the owners of residential properties rather than businesses, although the fact that businesses get more than half of the property tax exemptions granted by state and local jurisdictions offsets this bias somewhat.

Finally, it is certain that, over time, residential real-estate wealth has assumed an increasing portion of the burden of the property tax. Prior to Measures 5 and 50, owners of residential properties remitted less than 45 percent of all property tax payments; today they pay well over half.

However, it is by no means certain that this is due to Measures 5 and 50. It is entirely possible, for example, that the real cause is Oregon's system of land-use regulation, which has, over time, increased residential real estate values faster than it has commercial real estate values. One interpretation of the figure above is that the shift, which OCPP attributes to Measures 5 and 50, is simply the result of long-term trends in the relative growth and value of residential real estate, which would have occurred even if Measures 5 and 50 had never happened. This view is entirely consistent with the conventional wisdom, which holds that, by tying residential property taxes directly to booming real market values, Measure 5 inadvertently gave businesses (commercial, industrial property owners) a lot more tax relief than it gave homeowners, that Measure 50’s cuts were aimed at redressing this imbalance and, that, by rolling back residential property tax assessments and restraining assessment growth, they were largely successful.

Other property tax considerations relevant to IP-28

Most local jurisdictions have compensated for the loss of property-tax revenue wrought by Measures 5 and 50 by increasing user fees. Consequently, to the plurality of economists who see property taxes, at least those levied in support of local services (which may or may not include public schools), as user fees, this shift is largely a matter of complete indifference. Furthermore, most of the increased user fees are remitted by businesses, which suggests that, even if Measures 5 and 50 had had the effect of shifting a portion of the burden of property taxes from businesses to residences, user fee increases probably more than made up for the difference (although, as is the case with property taxes, the ultimate incidence of certain of those user fees remains somewhat unsettled, so that is by no means necessarily the case).

There are two important caveats that should be raised here. First, these conclusions do not apply to public schools. Instead, Measure 5 shifted responsibility for school funding from local districts to the state, which has failed to fully offset the effects of Measures 5 and 50, and, over the past 15 years, the state has allowed things to get progressively worse. Second, the state has imposed property-tax rate caps on local tax districts in a dozen or so counties that are not subject to Measure 5 compression (i.e., have combined statutory rates of less than 1.5 percent). Those caps have caused a lot of unnecessary hardship and are stupid.

[1]I calculated an effective tax rate for industrial and business and commercial property as well, 1.41 percent. However, for that purpose I used a different data set, from the census rather than from the DoR, which may or may not be consistent with the figures reported above.

Monday, July 18, 2016

More Background on IP28 (Measure 97?)

Fred Thompson checks in with another post on IP28. 

Over the next 3 months, proponents and opponents of this Measure will make a lot of wild, unsubstantiated claims about what it will or won’t do. My purpose it to try to distinguish sense from spin, as honestly and objectively as I can. Frankly, this is not easy. Not only because IP28 is unlike anything found in the public finance universe and is, therefore, extremely hard to parse, but also because it appears to be nearly everything a good tax is not. Tax experts believe that good taxes should treat the things being taxed more or less equally, that they should be fair, broad based, transparent in incidence and effects, and straightforward in administration. IP28 is unequal, probably regressive, narrow, opaque, and likely to encourage further avoidance. Consequently, every fiber of my being as student of public finance wants to say it’s spinach, to Hell with it. But I’ll do the best I can, starting with claims made about the existing tax system.

Proponents of IP28 claim that “Oregon has the lowest corporate taxes in the country.” What that means to an economist is that Oregon ranks dead last in corporate income tax (CIT) collections or that the growth of its CIT collections have lagged those of all other states.[1] Clearly, neither of those claims is true.

This figure shows the composition of each state’s tax portfolio in 2008 by tax type. CITes are represented in powder blue, personal income taxes (PIT) in light pink. Oregon’s CIT receipts as proportion of disposable income ranks 25th out of 50 states – right smack dab in the middle of the pack.[2] The most distinctive thing about OR compared with other states is its heavy reliance on the PIT – the fairest and most transparent of the tax types used widely by state governments.

Nor is it true that Oregon has a “very low corporate tax rate.” Oregon’s statutory CIT rate is above the national average. Nevertheless, there is no question that CIT collections have lagged the growth in business profits over the past 40 years. That is true at the national level, it’s especially true at the state level – and it’s true in Oregon as well. But it isn’t true that once upon a time Oregon relied heavily on the CIT or that its decline has impoverished Oregon’s state government.

In Figure 8e, Oregon’s CIT is shown in light pink. The key takeaway from this figure is that Oregon’s CIT never really played a dominant role in Oregon’s tax system. Other revenue sources have always overwhelmed the CIT, both in terms of totals and even annual fluctuations. The one thing that this figure doesn’t clearly show is the decline of the CIT relative to other revenue sources. Figure 8g shows the composition of Oregon’s revenue portfolio over time. Again the CIT is depicted by the light pink and it clearly tells a story of relative decline – that, relative to Oregon’s portfolio of revenue sources, the CIT is only about half as important as it once was. Figure 8g also shows that weight of ‘other revenue’ and severance taxes in the state’s revenue system has fallen even faster.

The Oregon Center for Public Policy (OCPP) recently reported on the causes of the relative decline in Oregon’s CIT. Their analysis is entirely accurate, although puzzlingly they list the causes in reverse order of their importance. In a previous blog I attributed Oregon’s CIT base erosion to the following: 58 percent due to the U.S. tax code’s expanded pass-through provisions, 31 percent to corporate tax sheltering and planning, and 11 percent to state-specific deductions, exemptions, and credits. Ironically, IP28 directly affects C-corps, the only business class that cannot avail itself of the U.S. tax code’s pass-through provisions. It is also probably the case, although OCPP denies the possibility, that a substantial portion (1/2-2/3) of the CIT receipts lost to pass-through is recouped in terms of higher PIT receipts.

Ultimately, however, Our Oregon’s claim that Oregon’s C-corps are not paying their fair share of taxes isn’t based on state CITs or even the tax liabilities specific to C-corps. Rather their claims are based on the Anderson Economic Group’s (AEG) 2016 State Business Tax Burden Rankings report, which ranks states according to the remittances of ALL businesses to state and local governments as a percentage of their pre-tax gross operating margins – i.e., the sum of the Bureau of Economic Analysis’s gross operating surplus measure plus ALL state and local sales, excise, severance, property, corporate-income, and business-income tax payments, license fees, and unemployment-insurance premiums.

That’s a silly gauge, both from the standpoint of the numerator and the denominator. Mostly, what the AEG report says is that the state of Oregon relies heavily on the PIT, which is IMHO a feature, not a bug. Indeed, not all the items included in the AEG report are taxes; some are fees for services rendered (UI premiums), but AEG appears to arbitrarily include some charges and exclude others (most local imposts, for example; these are relevant to us because Oregon’s are the highest) in the US. More significantly, tax burdens don’t depend on who writes the checks but on who suffers the consequent loss of buying power. Some of the taxes remitted by businesses can be reasonably assumed to stick with their owners (business income and corporate income taxes, mostly; property taxes, probably); the incidence of the other taxes they remit is largely shifted forward to final consumers (sales and gross receipt taxes) or backward to employees (UI premiums) or other factor suppliers (severance taxes).

One good thing about the AEG report is that it looks beyond state government, recognizing that local taxes are part of the state system, even if they do so in what I think is a wrong-headed and haphazard manner. In a future blog, I’ll take a look at the claim made by supporters of IP28 that, as a result of Measures 5 and 50, businesses in general and C-corps in particular are ducking their fair share of Oregon property taxes.

[1] They further say that Oregon “small businesses pay 8 times the tax rate of big corporations.” Their reported source is the Department of Revenue, but I cannot substantiate this claim one-way or the other. It is clearly not the case if one looks only at the state’s CIT and business excise tax.  They also claim that “[c]orporations with more than $25 million in Oregon sales paying minimum taxes pay the equivalent of only 3 cents for every $100 in sales,” citing the same source. Again, I can neither confirm nor deny this claim, but it sounds credible, assuming that we are talking only about the state’s CIT and business excise tax.

[2] This would also appear to put paid to the claim that Oregon taxes corporations at an unusually low rate “due in large part to the fact that most corporations (72%) pay just the minimum, and those minimums are very low." In fact, Oregon’s alternative minimum tax is the exception, not the rule. This claim looks like question begging, pure and simple.