Saturday, November 8, 2014

Medical Device Excise Tax and Transfer Pricing

The Republicans ran on tax cuts for the rich – well the “job creators” even if getting tax breaks will not necessarily led to any new job creation. But it is time to reward the base:
First on the chopping block is likely the 2.3 percent tax on medical devices, such as hospital beds, MRIs, pacemakers or artificial joints. Soon-to-be Senate Majority Leader Mitch McConnell of Kentucky immediately targeted the tax in his victory speech Wednesday. The tax has been in place since 2013, and medical device lobbying groups have spent millions trying to gain support for repeal
Jane G. Gravelle and Sean Lowry of the Congressional Research Service just provided an economic analysis of the Medical Device Excise Tax (MDET), which may provide some ammunition for the proponents of repeal. Their report turns out to be a fairly balanced presentation of the issues. Page 8 points out something that needed further development:
A July 2014 report issued by the Treasury Inspector General for Tax Administration (TIGTA) found that the number of medical device excise tax filings and the amount of associated revenue reported are lower than estimated … The IRS estimated between 9,000 and 15,600 quarterly Form 720 tax returns with excise tax revenue of $1.2 billion for this same, two-quarter period. In other words, actual medical device tax collections were 76.1% of projected collections during this period.
Paul Jenks echoes this shortfall:
Through the first half of 2013, Treasury auditors estimate that the tax levy should have collected $1.2 billion in excise taxes, but the IRS has received $913 million.
Was the problem non-compliance by firms who did not realize they were covered or was there something else going on? What is missing is any discussion of the transfer pricing aspects. Richard Ainsworth, Andrew Shact, and Gail Wasylyshyn (ASW) provides a nice discussion of the constructive price issue:
Related party pricing - Will the traditional constructive price rules apply in the medical devices area, most notably the 75% valuation safe harbor rule under Revenue Ruling 80-273? Given the related party transactions typical in the medical device market and the probability that transactions may be structured among related parties to reduce exposure to the MDET, this issue could impact the revenue raised under this tax …The proposed regulations indicate that there should be a “… basic sales price [that] assume[s] that the manufacturer sells the taxable article in an arm’s length transaction (that is, in a transaction between two unrelated parties) to a wholesale distributor that then sells the taxable article to a retailer that resells to consumers.” The basic sales price presumes a traditional manufacturer-wholesaler-retailer-consumer marketplace. In cases where the basic sales price is not available, a constructive price will be determined, as an exception .. The medical device industry is exceedingly top-heavy. Although the Medical Device Manufacturers Association (MDMA) observes that 80% to 98% of the medical device manufacturers swept up in the MDET will be small businesses, 86% of the $20 billion the MDET is expected to generate over ten years will come from the ten largest firms. Thus, even though the MDET will be a considerable administrative burden for numerous companies, the government’s revenue will come primarily from the largest publicly traded multi-entity groups.
What does this all mean in terms of how much tax will be collected by MDET. Let’s assume a medical device manufacturer with $10 billion in U.S. sales per year. I’m thinking of this firm or the medical device division of this firm. MDET will not cost our firm $230 million at all. At most the tax bite would be only $172 million per year assuming these firms adopted the 75% safe harbor. But they are most likely paying less if not a lot less. I will argue shortly that the constructive price should more likely be only 65% of the end user sales price, which would mean our firm would be $150 million per year in MDET. The reality, however, is that they are only paying $69 million per year through a simple form of transfer pricing manipulation. The Big Four issued commentary on this issue similar to the following:
In the absence of further IRS guidance, taxpayers subject to the MDET who are concerned that the safe harbors set too high a constructive sales price or who have transaction patterns falling outside the safe harbors may decide to take approaches similar to those used elsewhere for U.S. federal income tax purposes to establish arm’s length prices. The taxpayer should first determine if it has any sales of a particular medical device at arm’s length at the same point in the distribution chain that can be used as a proxy to determine the constructive sale price for transactions with related parties involving such device. If not, taxpayers may be able to use certain transactional transfer pricing methods to support a more realistic constructive sale price for MDET purposes. The IRS has extensive guidance under the transfer pricing rules of section 482 on the “arm’s length principle” and permissible methods for establishing appropriate transfer pricing for U.S. federal income tax purposes. Very generally, the arm’s length principle of section 482 (including its various acceptable transfer pricing methodologies) is intended to produce a price that is consistent with the results of a transaction “if uncontrolled taxpayers had engaged in the same transaction under the same circumstances (arm’s length result).” Although the Treasury Department and the IRS made a curious comment in the preamble to the final MDET regulations suggesting that transfer pricing methods for section 482 purposes are not appropriate for MDET purposes, taxpayers may be skeptical about the statement’s significance given that constructive sales price rules and the transfer pricing rules generally have the same objective. Similar to the arm’s length principle articulated in section 482, the constructive sales price for purposes of a manufacturer’s excise tax (including the MDET) is “the price for which such articles are sold, in the ordinary course of trade, by manufacturers or producers thereof, as determined by the Secretary.”
That comment that section 482 principles are not appropriate for MDET purposes is indeed curious. I would argue that based on an appropriate application of the Resale Price Method that one could readily argue for a discount between 30% and 35%. The 30% argument comes from the observed gross margin of companies such as this one or this one. ASW also notes that if there is a service division as well as a sales division for a medical device manufacturer, their value-added should be excluded from the constructive price. Let’s assume these divisions combined operating expenses are 30% of sales and an appropriate operating margin is 5%. Then you get my 35% discount. So how would the medical device manufacturers pay only $69 million and not my $150 million? Because the Big Four accounting firms are arguing for discounts that are twice my answer. How on earth do they justify this extreme result? It is called the Cost Plus Method with production costs being 25% of sales and a contract manufacturer return equal to 5% of sales. Of course, the $3.5 billion difference between the Big Four answer and the 35% discount rate under the Resale Price Method represents the value of the product intangibles of the medical device manufacturer. Under arm’s length pricing, no manufacturer would fail to include this amount in their price to a distributor. So how are the Big Four writing these reports with a straight face? The answer is simple – they are advocacy reports based on the assumption that the IRS is stupid.

1 comment:

Thomas L. Hutcheson said...

If the tax is not replaced by another tax or by reduced spending, repeal will be a good thing. It has never been clear what externality it was directed at and was probably dictated by the mistaken desire to reduce the fiscal "cost" of ACA. When the time comes to increase taxes, a progressive tax on consumption is a better option.