13 February 2019

Taxation is no way to spur innovation


When you tax something, you get less of it. That’s why when governments want to raise revenue to fund public services, they need to pay close attention to how taxes affect behaviour.

In general, they would be wise to tax ‘bads’ (e.g. tobacco, pollution, and congestion) before taxing ‘goods’ (e.g. work, investment, and R&D). If you are forced to tax a ‘good’, then you are best off targeting the ‘goods’ that are unresponsive or, to use the terminology of an economist, inelastic to changes in prices (e.g. land).

Yet an increasingly influential idea in innovation policy argues the opposite. Its advocates, who include Mariana Mazzucato and William Lazonick, argue that companies like Google and Tesla are to an extent free-riders because they benefit from taxpayer-funded research. In their view, the state takes the real risks funding research with highly uncertain payoffs, while industry gets all the rewards from commercialisation.

In a recent article, Mazzucato and Lazonick mention the possibility of using equity stakes to ‘pay it forward’. Using the example of the National Science Foundation (NSF) funding the research that developed the PageRank algorithm, they state: “Google could give back a return to the National Science Foundation that funded its original algorithm so that the NSF can fund more Googles.”

Take a recent line of questioning from Rep. Alexandra Ocasio-Cortez to Aaron Kesselheim of Harvard on National Institute of Health spending:

AOC: “Would it be correct, Dr. Kesselheim, to characterize the NIH money that is being used in development and research as an early investment? So the public is acting as an early investor in the production of these drugs. Is the public receiving any sort of direct return on that investment from the highly profitable drugs that are developed from that research?”

AK: “No, in most cases there is – when those products are eventually handed off to a for-profit company, there aren’t licensing deals that bring money back into the coffers of the NIH. That usually doesn’t happen.”

AOC: “So the public is acting as an early investor, putting tons of money in the development of drugs that then become privatised, and then they receive no return on the investment that they have made.”

The implication is that the state ought to take a stake in businesses that commercialise publicly-funded research. Indeed, the idea features in the recently proposed Green New Deal.

Taxes, you’ve invented taxes

At this point you might ask ‘don’t we already do this with taxes?’ The answer is, well, kinda. The main distinction is to levy higher taxes on companies that have benefited from public goods. As Stian Westlake points out: “The end result of this would be a fee-for-service public sector”.

There might be a case for a fee-for-service public sector in in some areas. For instance, moving to a system of use-based charging for roads and parking would both encourage road users to switch to more efficient forms of transport (e.g. buses and trains) and help forecast demand for future road improvements. But I don’t think this logic makes sense for all public services.

While some goods are congestable – scientific knowledge isn’t. If you drive on a road, use energy to heat your home, or book a table at a restaurant, there’s less stuff to go round for everyone else. Ideas are different: they are nonrival. You don’t deplete the stock of ideas when you use an idea to develop a new production process. In fact, you do the opposite. Knowledge tends to spill over. Even if I’ve patented my new technique, the fact I’ve brought the idea out into the open will lead to follow-on innovation.

Why innovation matters

It’s hard for an innovator to fully capture the benefits from producing a new idea. Most of the benefits of innovation flow to consumers, not to the innovative business. When Sanofi or Apple develop a new life-saving drug or smartphone, consumers benefit massively. In fact, William Nordhaus calculated that innovative firms only capture 2 per cent of the total benefits of innovation.

The benefits to society from innovation are so large that public policy should be biased towards figuring out ways to accelerate the pace of innovation rather than divide up the returns from innovation more evenly. A 1.2 per cent increase in the stock of innovation would provide the same benefit to the rest of society as a 50 per cent appropriation of the benefits captured by innovative companies.

Bringing new ideas to market, even for firms with minimal technical risk, is tricky. Even when academics have identified a promising protein or cellular pathway, the path to producing a marketable drug is fraught. It is not as simple as swooping in after the hard stuff has been done and then funding the easy but expensive trials (see drug researcher Derek Lowe’s excellent blog for an example).

The process is not only expensive but also risky. Drug companies are only willing to go to the effort of finding a drug lead and testing for efficacy and safety if they’re convinced they’ll get a decent upside. Blockbuster drugs are rare and the large pay-offs cover the costs of the flops.

Now suppose the public bodies who funded the initial research were to take a stake in the companies who brought the drug to market (let’s ignore the administrative challenges for the time being). To at least some extent, this would reduce the incentive for drug companies to partake in the risky process of bringing new drugs to market. Society may have a greater share of the benefits from a new drug, but they would have fewer new drugs.

Suppose instead the government developed a new open source productivity tool. Would society be better off if we forced any company that used it to give up equity? I’d suggest not. Charging may have other benefits, such as signalling demand for the product, and may be useful for determining whether or not to update and maintain the software, but if the sole aim is to maximise society’s benefits from innovation then it’s hard to argue against free distribution.

But aside from hypothecation, it’s hard to see the proposed system leading to more innovation than the status quo. At the moment, basic scientific research is funded from a broad tax base. Mazzucato and Lazonick’s alternative would supplement that funding with an additional tax on a narrow base (firms that benefit from publicly funded research).

In other words, they are advocating a system that directly taxes innovators more for the purpose of promoting innovation. If incentives matter at all, this won’t work.

CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.

Sam Dumitriu is Research Director at The Entrepreneurs Network.