New research shows the downsides of scrapping tax havens

Generally analyzed through the lens of tax collection, a country’s creation of a tax haven within its borders is universally deemed lousy policy.

Domestic firms use them to shift profits to affiliates where a lower rate applies and thus reduce their company-wide effective corporate tax rate, with the new job and investment gains concentrated in the lower-rate jurisdiction generally not enough to make up for the loss in revenue from profit-shifting.

{mosads}However, a working paper by Juan Carlos Suárez Serrato of Duke may be about to spark afterthoughts about stylized models like these that have long informed efforts to tame profit-shifting in successive reforms of our corporate tax code.

Using as a natural experiment the terminal phase-out in 2006 of corporate tax exemptions to affiliates of U.S. companies setting shop in Puerto Rico, the research finds that scrapping the island’s status as a tax haven led U.S. companies to cut back investments and job creation in the mainland substantially.

The provision in question was Section 936 of the Internal Revenue Code and it exempted Puerto Rico-based affiliates of U.S. companies from paying any corporate income tax altogether. At the time of repeal in 2006, §936’s value to exposed companies reached $232 billion in revenue, amounting to 15 percent of total profits shifted out of the U.S. that year.

After succeeding in boosting living standards on the island far above those of regional competing economies since its rollout in 1976, opposition to §936 spanned Congressional majorities on the mainland and Puerto Rico’s pro-statehood government through the 1990s. In 1996, President Clinton signed the Small Business Job Creation Act, spelling §936’s full phaseout by 2006.

With extensive macro data from the 15 years since then and sound methods to control for exogenous factors unrelated to the specific change in tax law, the research finds sizable harm inflicted by §936’s repeal country-wide.

Though Puerto Rico’s share of U.S. GDP was less than 0.63 percent at the time of repeal, ditching §936 appears to have raised U.S. companies’ average effective tax rate on domestic corporate income by 10 percentage points.

Notably yet unsurprisingly, they responded by cutting global investment by a whopping 23 percent while balancing away from domestic projects, in Puerto Rico and the mainland alike — domestic investment fell by 38 percent, with Foreign Direct Investments’ (FDIs) share of the total growing 17.5 percent.

As happens often with large employers such as US multinationals with affiliates in other jurisdictions, the global investment picture translates predictably to jobs at home. Employing 11 million workers in the continental US before repeal, firms taking advantage of §936 laid off a million of them, amounting to a 9.1 percent decline in payrolls.

The paper’s second set of results focuses on the concentrated effects of the aforementioned job and investment losses in those mainland counties where §936-reliant companies were most active. To this day, wages, employment and home values are all rising slower in those areas than elsewhere, and metrics such as the aggregate value of unemployment benefits or income replacement programs are conversely falling slower.

Mainland counties at the 75th percentile of the §936 exposure distribution — in other words, those with only 25 percent of other counties showing higher employment numbers by §936-reliant firms — have seen jobs growth 7.22 percentage points slower than counties at the 25th percentile. Similarly, moving from the former percentile to the latter results in income growth 12.55 percentage points higher.

In terms of unemployment benefits and income replacement, counties in the more affected group of the two received an additional $16 dollars per capita of the former and $30 of the latter than the other group of counties. Similar results apply for house values and rent, with the brunt of the hit borne by non-college educated workers across all the variables.

The paper’s findings beg some further rethinking of the standard rationale behind the establishment of tax havens. Despite uncontested success in bridging some of the gaps in manufacturing activity and other economic fundamentals between the mainland and Puerto Rico through the three decades it was in place, was §936 wise policy in the first place given the sizable job and investment losses from scrapping it far beyond the island that we know of today?

But more importantly, the analysis casts doubt over the general agreeability of the minimum intangible tax language in last year’s tax reform bill. One of four so-called “anti-erosion” provisions paired to the transition towards a territorial system for taxing profits, the Tax Cuts and Jobs Act defines Global Intangible Low Tax Income (GILTI) as a whole new category of corporate income to allow for foreign returns on a company’s assets above 10 percent to be taxed at a worldwide minimum rate between 10.5 percent and 13.125 percent.

By slapping taxes on sizeable profits from foreign intangible assets such as patents, the GILTI seeks to tame the incentive to shift those profits abroad through the use of intellectual property (IP). However, companies that have so far benefited from the absence of such a minimum rate threshold on foreign-held IP proceeds will likely pass on their hiked tax bill to domestic consumers and workers in the form of less job creation, slower wage growth or even higher prices — similarly to the §936-reliant companies on Serrato’s dataset.

To be sure, §936 shows that desperate times call for desperate measures even when it comes to engaging in academically vetoed tax policy. In the strictest sense and despite the recent enormous setbacks stemming from two back-to-back category five hurricanes, the provision fulfilled its promise of raising economic standards in Puerto Rico above those of comparable Caribbean economies.

But going forward, similar efforts to curb base erosion by standardizing the incentives inherent in the tax code may run into the same bottom line lesson to be drawn from the repeal of §936: never set up a tax haven, but once it’s on, better not scrap it.

Jorge González-Gallarza is a policy associate at E21, the nonprofit Manhattan Institute‘s economic policy shop in Washington, D.C.