Not only is the former host of the US version of The Apprentice now the country’s president, but even business location decisions there sometimes resemble reality TV shows. Amazon’s fourteen-month beauty contest to find a host city for its second US headquarters (HQ2) concluded in high drama this week. The company eventually opted to split the proposed 50,000-job venture between Queens, New York, and the Virginia suburbs of Washington DC. A consolation prize of an operations centre was granted to Nashville, Tennessee.
This three-way division was a fitting end to a sorry exhibition of crony capitalism. But perhaps there’s a silver lining: a reduced likelihood that cities, regions, and even the UK, will engage in such company-specific favouritism in future.
America’s system of federalism and localism is, by and large, a source of economic strength. It enables experimentation which enhances dynamism. Competition between states and cities pressures legislators to maintain broad-based business-friendly environments. People and business can vote with their feet.
As perhaps the most centralised country in the developed world, the UK could learn a lot from it. OECD data shows here just 5pc of total tax revenue is raised locally, compared with 33pc at the level of state and local government in America. The UK’s absence of meaningful regional economic policy levers can condemn poor areas to struggle for decades.
Yet an evident downside of the US model occurs when massive companies announce intentions for a new headquarters or plant. Cities compete in an economically destructive way to attract them. They stumble over themselves with cronyist incentives, on everything from company-specific tax incentives and infrastructure promises right through to job-training programmes.
In Amazon’s case, 238 cities initially entered the “contest” to host HQ2. The company, having outlined the public goods requirements it sought, then whittled this down to 20 “semi-finalists”. Pretty soon the process, in effect, became a bidding war with taxpayer money. The Mercatus Centre estimated that the average “offers” of subsidies and tax breaks of the semi-finalists were $2.15bn (£1.6bn) from cities and $6.75bn from states over a 15-year period.
Offering such incentives is understandably appealing to politicians. Securing a headquarters or plant, with the promise of jobs paying an average salary of $150,000 and up to $5bn of investment, is a visible win that can be seen to transform an area. The lawmakers might even get a prominent role when the ribbon is cut. On crude fiscal calculations too, if the uptick in tax revenues from the wages and spillovers exceeds the tax breaks and subsidies, legislators can claim they are improving the public finances, on net.
But this ignores the broader costs. The companies that might have invested if the state or city had improved the business environment more broadly. The corruption. The waste. Academic research concludes that development subsidies don’t improve overall growth against comparable cities or regions. Politicians are bad at backing winners. But even if it were otherwise, at a national level such incentives and the whole charade of the contest certainly reduce economic welfare.
In the Amazon case, 235 “losing” cities expended hundreds of thousands of dollars putting bids together, ultimately fruitlessly. The company eventually chose cities with large pools of skilled employees and other desirable features, despite more stingy incentives than offered elsewhere. Yes, these less generous packages are reflective of New York and Northern Virginia facing larger costs than some under-utilised cities: the pair have tighter housing markets, and face more physical congestion in transport, adding capacity to which will require significant investment. But we must surely conclude that by playing off the cities against each other, Amazon extracted monetary incentives even from cities they were likely to opt for anyway.
The direct tax breaks and subsidies are not massive. In Northern Virginia, they amounted to $25,000 per job, or $2,000 per job per year for the time the incentives imply. For New York, it is higher, at $6,100 per job per year. Some deals of this kind are better thought of as useful commitment devices to prevent future governments hitting major companies unfairly. Amazon’s current headquartering city, Seattle, recently proposed a head tax on tech companies, for example. This amounted to a crude attempt to extract more resources from the firm, and the company are keen to avoid the same.
But that doesn’t forgive the inherent unfairness to other firms not benefiting from the same treatment. More than that, it’s mystifying that Amazon, a company that evidently expects its investments to increase its sales and profits, will benefit upfront from such taxpayer largesse.
This is all important to remember post-Brexit. As trading arrangements change, there will be significant relocations of company headquarters and plants for the purposes of tax and serving markets. Threats about where to locate will be accompanied by demands for taxpayer support, probably through disguised forms of tax-breaks. Amazon’s example shows that the government should focus on a favourable regime for all businesses.
Indeed, the Amazon saga is also useful in thinking about future efforts to decentralise power within the UK. State aid rules have been relatively successful at the EU level from preventing arbitrary government favouritism. But if the UK were to decentralise taxes to regions internally, then firmer pro-competition legislation might be necessary to ensure that targeted incentives didn’t proliferate.
Amazon’s ultimate decision shows that other factors dominated modest firm-specific grants or tax breaks for them in choosing where to locate. The UK’s regions need far more economic policy autonomy to improve their attractiveness and structural dynamism. But throwing bungs at specific companies means that, over time, we all lose.