How did France's Robin Hood Tax work out?

Back in 2011 and 2012 I wrote quite a bit about the financial transaction tax (FTT) that was being proposed by the EU, and which is now being proposed by Labour. We have a FTT on shares in Britain in the form of stamp duty, but now Labour want to impose it on other financial assets too.

An FTT is a tax, usually a fraction of a percent, on trades of financial assets. It’s sometimes referred to as a Tobin Tax and is intended by its advocates to reduce high-frequency trading which, they claim, adds random noise to market prices and increases volatility. 

In this sense the tax may be efficiency-raising in that it shifts costs these traders impose on others onto those traders themselves. However, if high-frequency trading is not mostly based on randomness, as Sam Dumitriu argues here, then curbing it will slow down the incorporation of new information to market prices, increasing market volatility and making markets less efficient.

FTTs are also sometimes referred to as ‘Robin Hood Taxes’ by people who think they will raise lots of money. Some of these people look at the total volume of a market and presume that taking 0.5% of that will not reduce the volume of trading very much. This is somewhat at odds with the other case for the tax – if it changes behaviour it won’t raise much money, if it raises much money it won’t reduce the supposed negative externalities that high-frequency trading creates.

The EU still has not implemented a Eurozone-wide one. France, however, did introduce a 0.2% tax on purchases of shares in any publicly traded company with a market cap above €1bn (£789m), on “naked” short sales of sovereign credit default swaps, and on some high-frequency trading. It included several provisions that made it less restrictive than a true FTT, including a rebate scheme for trades of the most liquid shares and for intraday trading. Still, we can look at this to see what the impact of stamp duty on shares might be in Britain, and what extending this to other assets might be.

At the time I pointed to the empirical evidence and warned that France’s tax could reduce market liquidity and raise volatility, and not raise very much money either. Sweden introduced a financial transaction tax in the 1980s which ended up raising one-fortieth of what its backers had promised, drove many exchanges to London, and was abandoned after less than ten years. I reckoned that France would see many similar results.

Five years on, what have been the consequences of France’s transaction tax? A recent European Central Bank paper looked at the impacts on French equity markets, and for the most part it concludes that the outcomes of the tax were quite bad.

Market liquidity fell a lot, by 20% for affected stocks. Revenues were much lower than anticipated – an estimated €475m instead of the €1.6bn that had been predicted – “which once again points at an underestimation of the impact on revenue-generating market activity”. 

Importantly, volatility rose and price efficiency fell, even though the tax exempted intraday trading (allowing “short-term arbitrageurs to continue eliminating price inefficiencies quickly”). Though the overall effect was small (but statistically significant), the rebate scheme masked a large effect on volatility in shares that were exempt.

The paper’s authors conclude that the French FTT had an overall negative impact on market quality, where reducing market liquidity led to less efficient pricing of assets and more volatile price movements, and effectively priced high-frequency trading out of existence altogether in affected shares. “The decrease in volume associated with the FTT hurt liquidity and crowded out “useful” trades.” 

In order words, as I said in 2012:

“High-frequency trading allows markets to be highly sensitive to new information and to intermediate between buyers and sellers who may not be in the market at the same time. Stopping high-frequency trading would have the effect of making price shifts more sudden, unpredictable and large.” 

More volatile markets are riskier ones, so the cost of investing is higher and you get less of it. We don’t want that. All of this should make us treat a British FTT with extreme caution, and indeed make us consider abolishing stamp duty on shares instead of extending it to other parts of the market.

PS: I can’t resist pointing out my other prediction, about a man whose approval ratings eventually fell to 4% and whose party now seems to be on its deathbed: “Along with policies like the 75 per cent upper tax rate Hollande proposed during the election, the French may soon regret electing their blundering new President.”

The great bidding war

Many general elections turn into bidding wars, and this election is certainly no exception.  Instead of proposing measures to grow the economy to the benefit of most sections of society, some parties vie to see how much money they can take away from a smaller group in order to hand some of it out to buy the votes of a larger group.

They promise largesse and nominate tax increases that they say will pay for it all.  Unfortunately, they never seem to have heard of a dynamic, as opposed to a static, economy, and seem unaware that taxation changes behavior.  So far we have seen proposed increases in Corporation tax and income tax hikes for high earners.  These have been variously promised to fund untold billions of extra spending on education, health, social care, public sector pay, and a raft of other things besides.

Critics have pointed out that the sums do not add up, but it is worse than that.  The proposed increases in Corporate tax and high earner income tax will almost certainly both lose revenue.  The steady reduction in Corporation tax, down from 28 percent to 19 percent, has seen huge increases in the revenue it generates, while the reduction in top rate income tax has not only increased revenue, but greatly increased the proportion of total income tax paid by high earners.  The top 1 percent of earners now contribute 27.5 percent of all income tax paid, while nearly half of the population pay no income tax at all.

Corporation tax falls on shareholders to some extent, and on labour to a much larger extent.  Higher tax rates lower investment and activity, and in some cases lead to relocations beyond the reach of the tax man.  Higher taxes on income reduce incentives and expansion, and encourage people to move away, while discouraging others from moving in. 

Increases in both of these lead to falls in revenue, so far from promising who will receive the imagined gains, those proposing the increases should be telling us how they intend to finance them.

Another example of something we've long maintained to be true

In rather too much of what it tries to do government is counterproductive. Examples abound, raising the minimum wage to aid the incomes of the poor leads to some having no incomes at all. Germany's vast push away from fossil fuels has led to a rise in the country's consumption of coal, cracking down upon illegal drugs just raises the profit margins for those who continue to deal them.

There is a subset of this, which is that government often designs plans to encourage some activity. Those plans being either so absurdly complicated, or come with such side effects, that they diminish, not increase, the desired amount of whatever it is. And so it is with forestry in England: 

The figures mean that a Government pledge that 12 per cent level of woodland cover should be reached by 2060 is looking increasingly remote.

What joy, there is a plan. That plan includes paying subsidies to those who plant trees:

Landowners who wish to plant a forest must negotiate a "complex and bureaucratic" system in order to obtain a Government grant, the report said. 

Three agencies, the Forestry Commission, Natural England and the Rural Payments Agency administer the main grant available, the Countryside Stewardship Scheme. 

...

Witnesses told the committee that the application process was “tortuous”,“bureaucratic”,“ overly complex” and “not fit for purpose”.

Quite how a simple grant to stick acorns in the ground can become bureaucratically tortuous we're not sure. It doesn't need the usual panoply of diversity advisers, health and safety checks, inequality impact statements and all the rest, does it? They're not out there demanding disability access and racial balance are they? 

Our own experience of business tells us that people shouldn't apply for government grants for anything. Precisely because to do so is to get sucked into a morass of bureaucracy making the money on offer not worth the effort. It's nice to have another example to add to the portfolio but there's got to be a better way of getting things done than this, no?

Perhaps even that those who would like a few trees around for their children get on with the acorns in holes thing and the rest of us agree that it's their land and nowt to do with us? Nor our money?

A bigger NHS budget wouldn't solve the ransomware problem, no

Some to many NHS trusts have found their computers encrypted, locked and held to ransom. Pay $300 in Bitcoin to... or else. At which point the usual and entirely unsurprising insistences that if only the NHS had a bigger budget then this would not have happened.

And that's not in fact true. But, of course, it is being said:

The ransomware attack is all about the insufficient funding of the NHS

No, as ever, this is not about the amount of money, this is about how the money is spent:

“The problem is that the old IT systems were never designed to withstand the forces now ranged against them,” Moores said. “You may note that US hospitals haven’t been so badly hurt if only because they have the money to use more up-to-date systems rather than coax older systems to keep going.”

No. Again, it's how money is spent, not how much there is.

Yes, there's that interesting side issue that the NHS used to pay Microsoft for support on that outdated Windows XP and then decided they wouldn't. But that makes no difference here as the company didn't release the necessary update to anyone until this attack had already started. Why should they, they announced they wouldn't be supporting this any more some years back.

No, our problem is that the NHS is a government system, those American hospitals are not. And the one thing that governments really aren't good at is maintenance. Anyone with any experience of the Soviet block in its pomp knows this.

We've even tried lavishing fortunes on NHS IT and it didn't work well.  Some £12 billion spent and by some reports not a single usable line of code resulted - possibly an exaggeration but still. And IT spending in the NHS is increasing at a reasonable clip, ahead of inflation at least.

Computer security is maintenance by the way. You don't need ribbon cuttings on masses of shiny new kit, there are no grand projects to announce, no one does stand up in Parliament and detail how 10 web monkeys upgraded 500 PCs yesterday. Which is exactly why a government run system ends up splashing £12 billion on nothing and spends nothing on security maintenance.  

It's entirely reasonable that the NHS runs on an old operating system by the way. There is so much sector and function specific code out there that upgrading would both be hugely costly and really not worth it. But money does have to be spent upon security maintenance, that one thing that a politically driven and centralised system never is any good at, maintenance.

Something that puzzles us deeply about the Labour Party Manifesto

We are told that government should have much more money. OK, we don't agree, but that's politics for you. We are also told that companies should pay much more in tax. Well, apart from the obvious point that companies never do pay tax, the incidence is always upon the wallet of some live human being, OK again, that's politics. No one has ever said that electoral platforms have to make logical sense.

But we do think that we'd all sort of hope that such electoral platforms will add up. Which is where our puzzlement comes in. For the claim is that government will get much more money by raising the rate of corporation tax:

UK corporation tax receipts surged to a record high during the past financial year despite the main rate falling from 30 per cent in 2008 to 19 per cent today.

We're really just not sure how, if lower rates bring in more money, higher rates will also bring in more money.

Bit of a puzzler, eh? 

The problem with over arching regulation is that it concentrates error

Yet another example of why market competition beats bureaucratic regulation. For if we have such competition then some certain amount of the people doing it are going to be wrong. Wrong in what they're doing it, how they're doing it, what they think people want to have done - the very fact that many people are doing it in many different manners is simple enough proof of that.

True, there's always the possibility that different people want different things done in different ways but we mean something more basic here, before we get to that stage. And thus we could indeed ponder the idea that if all the clever people just told us what to do then we'd avoid the waste of that competition and that wrongness.

Except it doesn't actually work that way:

Sir James Dyson has won a shock victory in the European courts over Brussels rules which the company says unfairly penalised its vacuum cleaners.

An appeal in the European Court of Justice said a previous ruling from a lower court against Dyson had “distorted the facts” and “erred in law”.

The EU decided that vacuum cleaners should be tested for their energy efficiency. They then went on to determine how they should be so tested.

The billionaire entrepreneur argued that the tests are only relevant to vacuums with their dust bags empty and do not cover them when they are full, as they would typically be in normal use. A full vacuum would typically use more energy.

Dyson’s vacuums use a “cyclonic” design without a bag to collect dust and the company argued that the tests used by the EU unfairly disadvantaged its products.

Two years ago the company argued in the European general court that there was a way of testing conventional vacuums when the bag was full, so a comparison could be drawn between the two designs.

The European Court of Justice today upheld parts of Dyson’s appeal, backing the company’s claims that tests are available to measure a vacuum’s performance when full.

It also backed the British company’s by saying tests should “measure the performance of vacuum cleaners in conditions as close as possible to actual conditions of use”.

That there is error in a system is one of those unfortunate truths about any system containing human beings. What we want therefore is a system which sorts through the errors as fast as possible and eliminates them.

Which is really what a market does for us, it's an experimentation machine. This works, that doesn't, this does but no one wants it, nope, they want this but it's currently impossible and so on. Bureaucratic regulation simply concentrates the error. The entire continent of Europe has had the energy performance of vacuum cleaners wrongly reported to it for years now precisely and exactly because we used the bureaucratic method of determining the rules by which we do so.

We'd all be much better served by Which? and the equivalents across Europe devising their own standards and that way we'd be able, through that terribly wasteful market trial and error, be able to zero in on the testing standard that was actually useful.

The effect of Labour's corporation tax

The Labour party has proposed increasing corporation tax to 26%, from its current rate of 19%, and in contrast to the government’s proposed reduction to 17%.

That would leave the UK with a tax rate one and a half times the level the government is proposing, putting us around the middle of European corporate tax rates rather than near the bottom (although still the lowest in the G7).

Labour claims that this will raise around £20 billion to fund various spending commitments.

On a simple mathematical basis, that looks about right.  The Treasury estimates that a 1% increase in the corporation tax rate would raise about £2.3 billion.  Multiply that by Labour’s proposed 9% increase and allow for inflation until 2021 when they propose to implement it, and £20 billion looks reasonable.

The problem is that the economy and business do not remain static while politicians fiddle with the tax rates.  People and companies react to changes; and the bigger the change is, the more they react.

A 9% increase in tax rates therefore is highly unlikely to raise nine times as much as a 1% increase.

Higher rates, lower revenues

One thing that is increasingly well evidenced is that higher tax rates generally result in lower than expected revenues.

This is not just shown by the outcomes of previous tax changes, but is also what we would expect from people reacting to tax changes.  When tax rates are increased, the return on investment is reduced, so there will be less investment and therefore less growth.  If investors and entrepreneurs will see less of the rewards, there will be fewer new businesses started up, less investment in expanding existing businesses, and fewer international businesses deciding to locate in the UK.

The effect of this is difficult to quantify, but to get an idea we can look at the time when corporation tax was last at the 26% proposed by Labour.  That was in 2011/12, and the tax at that level raised around £41 billion[1].  Add on inflation and that would be around £44.5 billion today.  However in fact, with the rate cut to 20%[2], the tax revenues have now soared to almost £50 billion a year.[3]

Although the rate is now significantly lower than it was in 2011, the corporation tax collected is actually higher because the lower tax rates have encouraged companies to set up or expand, or to set up operations in the UK.

If that trend continues, the government’s proposed 17% might raise around £52.5 billion, £8 billion more than it raised when rates were last at Labour’s proposed 26% (all figures at today’s prices).

Assuming that trend works the same in reverse, the proposed rate increase to 26% would be expected to reduce corporation tax revenues by £8 billion, not increase them by £20 billion, as the UK becomes a less attractive place for business investment.

Who bears the pain of corporation tax?

This is not just an abstract matter of changes in an index of GDP.  Nor is it merely a question of whether investors see the value of their investments fall.  The effect will be real and wide-ranging, because any reduction in investment means fewer jobs, and less well-paid jobs, as companies reduce investment and try to cut costs in response to higher taxes.

The Institute for Fiscal Studies responded to Labour’s proposal by saying that “taxes are paid by people” and so corporations do not actually pay tax.  They have been criticised for that by supporters of higher taxes, with one saying the IFS is “so obviously factually wrong … companies are separate legal persons … only they can pay the corporation tax a company owes”.  However that criticism confuses the practicalities of “paying”, transferring money to the tax authority, with an economic concept of payment in the sense of bearing the burden of the tax.

The truth behind the IFS claim is that people, rather than companies, suffer the burden of corporation tax.  Either the company has less wealth (so the shareholder bears the burden), or the company increases its prices to keep its after-tax income the same (so the consumer suffers), or wages are reduced, staff are laid off and new staff are not hired, to cut costs and maintain the same after-tax income (so the workforce bears the pain).

In practice there is a combination of the three.  However, in an open, free economy, there is usually not much scope to increase prices (the business would become uncompetitive), and if shareholders see too much of a cut in their returns then they will invest elsewhere.  That means that the main pain of increasing the corporation tax rate falls on the workforce, as the company seeks to cut costs to maintain its after-tax profits.

And not just the company’s employees that lose out; some of the main losers are young people trying to find their first jobs, only to discover that few companies are hiring because they have cut back on their expansion plans in the face of higher taxes.

Pretty much all economists are agreed on this; the only question is how much of the burden of a corporation tax rise falls on workers rather than shareholders or customers.  However a major study by Oxford University’s Centre for Business Tax[4] concluded that a rise of £100 in corporation tax would reduce wages by £75, through a combination of lower wages and fewer jobs.  Some studies have found even higher tax burdens on the workforce, others lower, but the Oxford study is one of the largest.

If 75% of the burden of increasing the tax rate falls on the workforce, that means that Labour’s proposed £20 billion a year from extra corporation tax receipts would reduce wages by £15 billion a year.

With average private sector wages of just under £26,500[5], that cut in companies’ salary bills is equivalent to over 565,000 jobs.

Knock-on tax losses

If wages are reduced by £15 billion as companies react to higher corporation tax rates, that will also reduce the government’s income tax and national insurance receipts.

On that average private sector wage of £26,500, the government would expect to take around £7,720 through income tax and National Insurance.  That means the lost wages of £15 billion could see a fall of £4.4 billion in the Treasury’s revenue from employment taxes.

Double whammy

Would the fall in wages happen as well as the fall in corporation tax receipts?   Yes, it could, because they are two results of the same process.

As tax rates are increased and companies invest less, because the after-tax rewards are lower, two things happen; company profits are lower, so there is less to tax, and also there are fewer jobs and those that remain are less well paid.

The effect then is potentially disastrous for the Treasury; lower company profits to tax and less employment taxation.  Plus of course the additional welfare costs; benefits for those who are unemployed because of the reduced investment, and in some cases higher tax credits for those who are still in work but on a lower income.

With a potential loss in corporation tax receipts of £8 billion and lost employment taxes of £4.4 billion, the government could be looking at a potential loss of over £12 billion of tax revenues, if this policy were implemented, not to mention the huge financial, social and personal cost of a possible 565,000 people losing (or failing to find) jobs.

The fact is that, although very important to the company, profits are a tiny part of what companies do; far more important are the goods and services that they provide and the employment opportunities they create.  Over-taxing those profits, which may only be a few percent of turnover, risks losing all the other advantages.

Although taxing companies looks like a painless way for the government to raise money, it is far from that; the pain of lost taxes and lost opportunities can be large and widespread.

[1] That figure is based on HMRC data for tax revenues in the later part of 2011/12 and the early part of 2012/13, because corporation tax is mostly paid in the following year.

[2] It is now reduced further to 19%, but because of the timing of when corporation tax is due, the first tax payments under the 19% rate will not be due until the end of 2017.

[3] Source: HM Revenue & Customs – receipts.

[4] “The Direct Incidence of Corporate Income Tax on Wages”, Arulampalam, Devereux & Maffini, Oxford, 2009.

[5] National Statistics, “EARN02 – average weekly earnings by sector”, January 2017 (latest finalised data

Maybe there are just too many such graduate teachers?

It is possible to somewhat make fun of these Yale graduate students who are on hunger strike over the terms and conditions of their work. As indeed some other Yale students have, by setting up a barbeque just next to their hunger striking station. The strike itself seems a little weak too, there are reports that anyone who feels really hungry can leave, eat, then come back. It is not eating while hunger striking publicly, rather than not eating apparently.

However, they're right, they have identified something of a major problem:

Unfortunately, this describes little of today’s reality. I knew this going into graduate school, but I went for it anyway, because I wanted to do nothing more than teach English in college the way it was taught to me.

....

This means two things for people like me, who are graduate teachers. First, it means that universities like Yale depend on us to teach their students but don’t give us any of the job security that the professors have. Second, although Yale says I’m in training to become a professor one day, it’s likely I’ll never have the chance.

Instead, if you check back with me in five or ten years, you’ll probably find me among the many thousands of people who have a PhD but can barely hang on to a place in the middle class. I may not have health insurance through my job, much less paid vacation, an office or any control over my schedule. In fact, it’s likely I won’t even know where my paycheck will come from further out than four months. In academia, you’re now less secure the more experience you accumulate.

Out here in the real world we're really pretty sure that when an occupation leaves you with a paltry income this is the world's way of saying that you should go and do something else. This is true of the business making a loss, the farmer unable to survive without subsidies and yes, the would be worker facing derisory pay as a result of too many other people clamouring to do the same job.

We're even really certain that a professor or two over in the economics department will be able to explain this to people. This is just what happens with an over supply of labour.

But the important, and unsaid, point is why is there such an over supply of said labour? The answer being that there are simply too many graduate students in American universities.

It's worth noting that a PhD over there is not, as it is here, the production of a thesis, the addition to the sum of human knowledge by some small amount. It takes rather longer, up to 7 years, and includes a certain amount of learning how to be a professor. Thus these teaching duties which they undertake. It's not entirely true but it is largely so that an American PhD is the training program to be a university professor in the fullness of time.

Which is where the problem is. The universities are training many more such professors manque than they will ever be able to employ as actual professors. Not just many more but many times more. And thus this problem of those who have this very expensive training but who are unable to grasp the brass ring at the end of it.

The answer is for those graduate programs to shrink considerably. Perhaps some subjects more than others - there's not really a use for advanced degrees in critical studies, gender studies perhaps, outside future employment in academia. And if the training courses are producing more than can be absorbed by the universities then it's the training courses that should be cut, no?  

But back to our basic economics. The bad pay and conditions of these people is a symptom of an over supply of people with this training. The conditions themselves should thus dissuade people from following this non-career path. But at root the diagnosis is simple. There is an over supply, people should be going and doing something else.

Don't forget, absolutely everyone else on the planet has also had to face the idea that there's something they'd love to be doing but there's a woeful shortage of people willing to pay them to do it. Odd that a university system is able to teach that point to its students really.

Cryptocurrency is interesting again

For the last few years, the world of blockchains - i.e., the world of Bitcoin and the technologies derived from it - has been firmly divided into two camps. The first of these, which can broadly be termed "Enterprise blockchain," is a group of companies, including Intel, IBM, and my own firm, Monax, which apply and adapt the ideas of distributed computation which Bitcoin pioneered for enterprise use-cases.

The second of these is the world of the "coins" or cryptocurrencies, Bitcoin and its imitators, which are doing the same thing they have always done: providing a means of value storage and transfer which is uncensorable and beyond the reach of state regulations (which is why Bitcoin has long been the darling of many libertarians). 

By all indications, Bitcoin is booming, with the price of one coin reaching all-time highs of $1,800 this week. Equally interesting, however, is the fact that the "altcoin" ecosystem - that is, Bitcoin's imitators - have exploded in value as well, to the point where their value equals nearly half of the value of the ecosystem as a whole:

OK. But why?

First, a short primer: any blockchain's application architecture is fundamentally different from what most of us use on a daily basis as we access the web. Applications like Dropbox or work e-mail are run on a server somewhere; administrators of those servers are firmly in control of them and place limits on our access rights to their databases, such as storage limits, to ensure that the system's performance is not degraded from overuse. 

Bitcoin, by contrast, is an automatic peer-to-peer system. It runs itself, with the assistance of all of its users, and has no central locus of control. For this reason, a copy of every Bitcoin transaction ever made is stored on every single computer (usually consumer-grade PCs) running the Bitcoin application, and those copies contain all of the rules about who can write to Bitcoin's system - rules which need to be agreed by all the users in advance. The brilliance of Bitcoin's designer(s) is that they wrote the rules in such a way as to obviate the need for human overseers and sophisticated enterprise hardware to run the application.

However, lacking a central enterprise cluster or a human administrator, a paramount architectural concern for Bitcoin is staying usable by preventing the chain from becoming unwieldy in the hands of its (mainly consumer/retail) userbase. This is achieved mainly in two ways; first, only someone holding a Bitcoin is able to write to (transact on) the database. Second, Bitcoin enforces a size limit of one megabyte on the numbers of transactions it will accept in any ten-minute period - a figure that roughly works out to 750 bytes, or 3 transactions, per second. 

Bitcoin has hit this limit. As a result, users are now forced to pay increasingly high fees to use the Bitcoin network (paying transaction processors additional Bitcoins to prioritise their transactions among the many thousands that are queued in a backlog, termed the "mempool," and shown on the chart above).

There are two emerging solutions to this - one technical, one market-driven. There is little consensus about how to approach a technical solution, even among the small group of elite developers who are Bitcoin's most respected maintainers. 

What *is* more or less agreed is that simply increasing transactional throughput is not an option: Bitcoin's transaction history is currently well over 100 *gigabytes* in size, and doubling transactional capacity means it would grow twice as quickly, meaning that re-centralisation a la Dropbox would occur - Bitcoin would grow so large that it could only be run on enterprise infrastructure.

The other option is the market solution, which the chart above appears to show. The cryptocurrency market as a whole is interesting from an economic perspective in that it provides a perfectly transparent sandbox to see what happens when perfectly substitutable goods (Bitcoin clones) that accomplish the exact same thing (unregulated value transfer) in a fully automatic way (distributed state machines which require no human oversight) are placed in a position to compete. As far as an end-user of cryptocurrency software is concerned, whether a c-currency is $3000 in Dogecoin or $3000 in Bitcoin is immaterial; the shop 'round the corner prices its goods in USD/GBP/EUR, so as long as one coin or the other has sufficient liquidity to cash out, this means competition can occur on the basis of speed and transaction fees.  

Which gives the other coins, in their role as fast-followers with smaller size, newer designs and greater community flexibility and incentive to win market share, comparative advantages.

Summing up, what this means is that cryptocurrency is getting interesting again. Not as an investment product (too risky for my taste) or because the technology is sufficiently magical and brilliant that it will eliminate banking (it won’t), but because it shows that even Bitcoin - a machine, which runs itself, and is not operated by any one firm or person - is, reassuringly, forced to cope with the oldest problem in capitalism. 

Chiefly, adapt to the perennial gale of creative destruction, or lose market share and die.

Taking innovation seriously

Matt Ridley made an interesting proposal in his Times column last week. In the piece he discussed the idea of the Her Majesty's Government adopting the innovation principle, which he summarises as "examine every policy, plan or political strategy for the impact it could have on innovation, and if you find evidence that the policy is going to impede it, then drop, change or rethink the thing."

He suggested that the Conservatives ought to "pledge to set up an innovation commission, guided by the innovation principle, whose job is to examine policies and report on their likely impact on innovation?" The Adam Smith Institute typically recoils at the thought of creating yet another quango, but I think it's worth making an exception in this case. If there's one thing in need of more regulation the regulators themselves.

But what would actually change if politicians and regulators were forced to take innovation seriously? Ridley suggests we'd have cheaper, safer food thanks to genetically modified crops and fewer people would smoke as burdensome EU rules on vaping would never get past the innovation commission.

I think there are two other areas where taking innovation seriously would lead to fundamental regulatory changes.

1. The Gig Economy

Innovative start-ups like Uber, TaskRabbit and Deliveroo are creating an on-demand economy that enriches consumers and offers new flexible models of working. Take Uber drivers for example: they're overwhelmingly happy to trade off traditional employment rights for the ability to work when and how long they want. Yet recently an employment tribunal ruled that Uber drivers should be classified as workers and that Uber be forced to pay the minimum wage and offer sick pay. This not only disrupts Uber's entire business model, it's having a chilling effect on other gig economy start-ups who are in a sort of legal limbo.

If regulators and courts were forced to subscribe to a form of the innovation principle, then legally ambiguous innovative workplace arrangements would be be given the benefit of the doubt.

This is because there's an asymmetry between over and under regulation. In medicine, we tend to be more concerned with false negatives than false positives. Better to be over-cautious and provide unnecessary treatment than let a condition get worse. The reverse is true when it comes to regulation. Market forces tend to mitigate the effects of bad behaviour. A monopolist can only keep prices so high before inviting competition, an exploiter can only keep wages so low before they lose workers to other employment opportunities. The costs of failing to crack down on bad behaviour are static and limited.

That's not the case with false positives. Wrongly cracking down on Uber or Deliveroo doesn't simply harm Uber and Deliveroo's shareholders and customers, it also hurts consumers widely by deterring innovators from entering the market providing similar services. Consumers are not just missing out on Uber and Deliveroo but missing out on their unborn competitors. The costs of shooting first and asking questions later are dynamic and unlimited.

2. 'Tech monopolies'

If Theresa May keeps her word, Britain is set to leave the jurisdiction of the European Court of Justice. The ECJ has typically taken an over-active approach to enforcing competition, often confusing market share with market power. If Britain were to take innovation seriously we ought to take an error-cost approach to enforcing competition law.

Judge Easterbrook set out the error-cost approach in his paper 'The Limits of Antitrust'. He shared the insight that false positives were much more harmful than false negatives. Wrongly penalising pro-competitive behaviour is more harmful than failing to act against anti-competitive behaviour. Market forces tend to correct against the latter creating incentives for new entrants to disrupt monopolists.

He observed that errors of both kinds were frequent because distinguishing between pro-competitive and anti-competitive behaviour is intrinsically difficult. It's especially difficult when dealing with new products and business models. Firms often don't fully grasp the full competitive effects of their business practices in advance, rather they employ a trial and error process.

Take Google for example. In 2016 the EU made an (in my view baseless) antitrust complaint against Google relating to its android operating system. But, Google is an excellent example of a firm experimenting with different business models without a clear understanding of their effects on competition. AdSense is responsible for nearly one fifth of Google's total revenue but was famously produced during Google's '20 percent time' where employees are given free range to work on projects of their own choosing. If firms themselves often fail to understand the full implications of product innovation until much later on. How can we expect courts to understand the full competitive effects of innovative business practices?

That's especially the case when dealing with companies like Google where they produce multiple complementary products. Antitrust law expert Geoff Manne highlights the problem with the EU's complaint about Google preinstalling Android apps:

Of course, Google receives some benefit from pre-installing its apps: Doing so provides a chance for the company to realize some return on its massive investments in the Android ecosystem by promoting its own products. Although Google requires many Android device makers to pre-install Google Search and to set it as the default (but not exclusive) search provider on their phones, this is hardly a bad thing: Search helps finance the development of Google’s other (free) apps, as well as (free and open-source) Android itself.

Imagine what might happen if Google were prevented from requiring device makers to pre-install Google Search or the Chrome browser—the apps from which it actually earns substantial revenue—as a condition of pre-installing its other apps on Android devices. Google would likely then charge hardware makers licensing fees to pre-install apps like Gmail and YouTube, the cost of which would be passed along to consumers in the form of higher device prices. This would hardly be a net gain for consumers, given that they already enjoy unrestricted app choice at lower cost today.

Alternatively, Google could vertically integrate like Apple, exercising tighter control over the Android ecosystem. This would be no boon for competing app developers, who are attracted to Android’s openness—just ask Spotify what it thinks of Apple.

A hypothetical innovation commission ought to force regulators to give businesses in markets with high levels of innovation the leeway to experiment with new business models. One way to do that would be to require direct proof of an anticompetitive effect and not merely rely on theoretical harms. Another would be to grant legality per se to new product introductions to remove regulatory uncertainty for innovators.

Innovation massively matters, but often regulations can have chilling effects and deter welfare-enhancing changes. An innovation commission could tip the scales in favour of innovation and prevent employment courts and competition regulators from keeping new products from the market.