It's almost as if they've no clue what they're doing

Two months ago we had this:

The transfer of 1.6 million patient records from an NHS trust to Google’s artificial intelligence subsidiary was legally inappropriate, the Department of Health’s data guardian has said.

Google’s Deep Mind division and the Royal Free trust in north London agreed two years ago to collaborate in developing an app to diagnose acute kidney injuries in NHS patients. It works by immediately reviewing blood test results for signs of deterioration.

To develop the Streams app, the trust made available to Google patient data going back five years, including sensitive details of patients who had not been treated or tested for kidney injuries. 

Using NHS data to train an AI to make the NHS work better is terrible. Those damn yankees and their technical skills, who do the bastard septics think they are trying to improve that Wonder of the World that is our very own National Health Service? They'll be forcing the ICU patients to pay $50 for an aspirin soon enough what with their only for money health care systems!


NHS patients will receive pioneering treatments assessed by an artificial intelligence program under plans being developed by the government’s medicines regulator, The Times has learnt.

The system, a prototype of which will come online next year, is expected to markedly speed up the appraisal of cancer drugs and other therapies from months to a matter of weeks.

Ooooh! Using NHS data to train an AI to make the NHS work better is just lovely.

Scientists at Microsoft, University College London (UCL) and Cochrane, a non-profit group that compiles reports on medical evidence, are developing an AI that can weigh up new health studies for itself. 

Yea, even if the septics are involved.

It's as if the government doesn't know what it's doing, isn't it? And that cannot possibly be true about the NHS, an absurd thought. For we all know that it's so exquisitely managed that no system in the world comes near it.

For once we agree with Polly Toynbee

This is odd of course but stopped clocks and all that*. Polly Toynbee makes a useful and sensible point. The great adventure of expanding university access in the name of social mobility seems to be failing:

Now that a degree leaves students up to £69,000 in debt, with 70% never earning enough to pay it all back, maybe the surprise is that few school leavers have been deterred. A graduate still earns around 35% more than a non-graduate. But averages deceive. While the number from poorer backgrounds hasn’t fallen, dispiriting new research finds that even after gaining a degree from a good university, those from poorer backgrounds, without the connections or the money to take internships, fare worse in jobs.

The Paired Peers project, which followed a cohort of students from Bristol University and the University of the West of England, found that their family’s social class still counted most, whichever university they attended.

Sending 50% of the age cohort off to university in an attempt to shake up the social classes doesn't seem to work. OK, so, experiment failed, let's not do that any more, eh? Back to a more reasonable 10 to 15% of the cohort taking these academic courses, around and about the likely number who will benefit from academia rather than just a course, and that of course makes the problem of funding the system much more manageable.

In fact, Polly makes two sensible points:

The second great shock, which simply defies belief, is the 23% slump in the number applying for nursing places this year, the first year when nursing students pay full fees and lose their bursaries. That is despite spending half their training time providing useful service to the NHS. 

We made noises about this when it was first mooted, that nursing should become a graduate  profession. Disapproving noises as well. Not just because we're the sort of snobs we are but at least one of us has direct experience of people training under the old and the new systems. That new is not better than the old from direct observation.

Sadly, Polly won't manage to make the third and correct observation from this. Just as there can indeed be market failures so also can there be government ones. The entire joy of a market based system is that when a mistake is made those making it go bust and disappear from the scene. Government mistakes not so much - how difficult does anyone think it's going to be to reverse these two mistakes?

*It is left as an exercise for the reader whether we are or Polly is that ceased timepiece

Planning reform can be safe as houses

Over on Medium I've written up what I think is a politically-achievable plan for the Conservatives to get some real action on housing now that gives them something to campaign on at the next election.

In housing, the root problem is mostly the planning system restricting supply – not enough nice, big homes are being built, which keeps prices higher than they need to be across the board. You're not going to win by promising planning reform or anything like it — unlike rent controls, they don't sound good. But you might win if you can show that housing is becoming more affordable and more secure. 

I propose a three-pronged approach — allow densification within cities, and have it done on a bottom-up, street-by-street level instead of exclusively through massive new developments; let local councils capture some of the uplift in land values that comes when planning permission is granted to new developments; and introduce a 'long-hold' midway point between shorthold tenancies and leaseholds, which effectively confer ownership of the property:

"Private rents in the UK are some of the highest in the EU, and private rented households spend between 35% and 40% of their post-tax income on rent compared to a European average of 28%. This does not capture the second-order problem caused by expensive housing costs, which is that it is much harder to move to economically prosperous parts of the country where better jobs are, so people end up forgoing better jobs and salaries than they might otherwise get.

"Housing quality is also quite poor. New builds in England are some of the smallest in the developed world, and shared living areas are being turned into extra bedrooms in many rented properties, squeezing more people in. In 1996 54% of 16–34 year olds owned their own homes; now only 34% do. That’s a twenty percentage-point drop in twenty years. Over that period the number of renters in that age category has doubled from from 1.1 to 2.2m.

"Labour made this a major part of its election campaign. Economists nearly unanimously agree that rent controls do harm, but many voters do not realise the risks. Bans on lettings agency fees and making three-year tenancies the norm similarly sound appealing to people fed up with wheeler-dealer agents and having to find somewhere new to live every year.

"These are tangible policies that sound good on the doorstep. The Tory manifesto was vague on housing issues and offered no track record of improvement. The government’s housing policies were basically useless — they only seemed to be interested in getting people to own their own homes, but because they did little on the supply side, policies like Help to Buy mostly only raised prices and changed the distribution of who got houses, not increase the total number of homeowners."

Read the whole thing, and my previous piece on some other policies the Tories should be going for to win at the next election.

The case against the new corn laws

There was once a time when Peel’s abolition of the Corn Laws was regarded as an unqualified success, ushering in an era of free trade and prosperity hitherto unforeseen in human history. What was once taken as dictum, however, seems to be lost on the current generation of policy-makers. The Corn Laws, along with the general spirit of protectionism they represent, has once again become fashionable under the guise of the Common Agricultural Policy (CAP). Needless to say, there is no reason to suppose that a policy which was so disastrous in the 19th century, would be any less disastrous in the 21st. It is remarkable how similar, and how similarly conclusive, the arguments against agriculture subsidies remain. The case against this nefariously illiberal policy is made by simultaneously appealing to the needs of the consumer, the competitiveness of the producer, and the health of the international market more generally. The vote to leave the European Union on the 23rd of June offers the perfect opportunity for the government to repeal these new Corn Laws.

Subsidies are simply redistribution under another name – that of ‘protection’. By definition, subsidy requires the productive sectors to finance the unproductive sectors of the economy through general taxation. In the case of domestic agriculture, 1% of the employment market is financed by the remaining 99%. This is made all the more inefficient by the inequity of national contributions to the CAP, with the UK contributing £6 billion, while receiving half that amount in subsidy. The CAP does not even contain the one redeeming feature of redistribution – that of relieving the plight of the least fortunate. The CAP takes all the wealth of the whole of society (including societies poorest) and redistributes said wealth to a group of landowners, who predominantly belong to the upper middle classes. Moreover, the further effect of this subsidy is to inflate domestic food prices by artificially raising the barriers to entry for foreign imports. Regarding the discussion of the ‘cost of living crisis’ at the previous general election, removing domestic agricultural subsidies would be an effective way of pushing down the price of food without distorting the market.

The CAP does not even achieve what it purports to achieve: the goal of keeping the domestic agricultural sector strong. Rather, the effect of the CAP has been to starve the agricultural sector of much needed competition and free enterprise. Instead of farmers competing against one another in order to create the most efficient product in a free market, the sector is now make up of a rentier class, each competing for a subsidy from government (whether national or supranational). In 1984, the government of New Zealand gradually reduced all subsidies and import quotas on agriculture, with the process finally completed in 1990. The agricultural sector of New Zealand’s economy (which is far more important when relatively compared to the UK’s agricultural sector) consequently boomed, as fair and proper conditions were returned to the market.

Perhaps chief among the CAPs crimes is that it keeps developing countries poor, and in constant need of aid. Repealing the subsidy would enable imports, from African and South American countries in particular, to be sold at a competitive price in Britain. This would not only allow British consumers access to cheaper food, but increase the economic strength of developing nations. This would, in turn, allow western nations (the very nations which imposed the subsidy), to reduce their own foreign aid budgets towards these countries, enabling prosperity, and reducing embezzlement by corrupt officials in one fell swoop.

All that remains is for Theresa May’s government to emulate its Peelite forebears, by scrapping agriculture subsidies, without exception, over a gradual period of years. It may not be politically advantageous – it is in the nature of embedded interests to cause political trouble when their privileges are questioned – but the future benefits of such a policy would go a long way to cementing May’s legacy at a time when her premiership has yet to begin. Before the referendum on the 23rd of June, Paddy Ashdown attempted to boost the Remain campaign by threatening that leaving the EU would ‘open the door to cheap food world-wide’. Let’s take him at his word. 

It's time to privatise London's buses

London is home to some of Europe’s most congested roads. While the London Underground does a good job of speeding passengers across the city beneath the gridlocked streets, London’s buses provide what has to be one of the slowest and least convenient public transport systems anywhere in Europe.

However London’s bus network is often held up as one of the country’s finest. A walk down any major London street will usually involve passing several dozen red double deckers in the space of a few minutes. The TfL bus network covers every imaginable corner of the Greater London urban area with over 500 routes, theoretically making any destination in the city reachable for a travelcard or Oyster user.

It is often a revealing exercise, however, to look at how many passengers these routes are carrying. There have been proposals by Sadiq Khan recently to remove the several hundred buses an hour that use Oxford Street to make it a more pedestrian-friendly environment, but there’s a huge question around where these displaced routes will go. However, the question is never asked whether all these routes need to exist at all. Any amount of time spent watching the buses on Oxford Street will reveal that many of them are running close to empty.

Supporters of central planning and public ownership will often point to comprehensive network coverage as one of the advantages of having public transport managed by one authority. London is a textbook example of this; although the routes are run by private operators, they are specified and tendered by Transport for London, in stark contrast to the completely private commercial operation found virtually everywhere else in Britain.

Elsewhere in Britain, operators are commercially forced to respond to passengers’ needs. Intense services appear where demand is highest, while services disappear where demand is lowest. One consequence of this is that smaller communities lose their services, while another is that the routes that emerge are far more attractive to potential users switching from other modes of travel.

This isn’t the case in London. While the TfL network does succeed in providing a service to every community in the city, it fails in providing a service that responds in any way to demand. London’s bus routes are indirect, circuitous and painfully slow. The lack of speed is partly down to the traffic situation, but stops every few hundred yards and routes that rarely head towards their final destination are the main thing that makes a bus ride in London so frustrating. It also doesn’t help that the numerous buses themselves are a major contributor to the congestion problem.

It hasn’t always been this way. A limited amount of private competition was introduced in London in the 1990s, with operators like Grey Green bringing a splash of colour to the red monotony that otherwise prevails. Latterly, however, TfL has tightened its grip, creating a situation where private operators with ideas for new routes are prevented from starting commercial services within the city, even where demand for different routes is demonstrably present. The omnipotent authority has even clamped down on branding variations by its own tender operators, as though Metroline’s blue stripe or Arriva’s cream swish were somehow damaging the quality of service being offered.

Imagine for a moment what London buses would be like under completely private operation. Gone would be the slow red double decker carting fresh air around every backstreet it could find, taking the best part of an hour to cross a distance the tube can cover in five minutes.  In its place would emerge direct point-to-point services, picking up major destinations and responding directly to the needs of passengers. Competition on key corridors would drive up standards, while operators would bring in quality service brands like Stagecoach Gold and Arriva’s “Sapphire” that have proved so successful at bringing people out of their cars across the country.

Certainly, some less densely populated areas would lose out. High-income areas where bus usership is lower would see service cuts. That’s how supply and demand works. There are far better ways of providing connectivity to remote communities than a frequent empty bus service, wasting money and fuel on a vehicle that could be better deployed where that capacity is really needed. It isn’t realistic to expect every community to have frequent bus services, any more than it is to expect every community to have motorway-grade road connections.

But other communities could stand to benefit hugely. Take the likes of Camberwell, a gigantic hole in the rail and tube networks where buses are the backbone of public transport. Here, direct demand-responsive services could act as street-level extensions of the underground network, bringing huge improvements to one of London’s most poorly served districts.

The current state of London’s bus network is a sad reflection on the conflicted political ideologies that have shaped it since the 1950s. It’s sobering to realise that this is what remains of what was once one of Europe’s most impressive tram and trolleybus systems, destroyed as it was by a political drive to free up road space for the car. Now, again, the passenger is being left behind in the same spirit of political idealism.

Let’s move to a system that operates for the passenger, not the politician. 

Even the IFS is at it now

One of the things which puzzles people - we know this because people ask us - is how come there is all this shouting about austerity? Cash spending is up, spending as a portion of GDP is still, just, up over what Gordon Brown was spending pre-crash. So, err, what austerity?

The answer being that those complaining about it all are using a different measure. Well, obviously, they must be, if their measure is not according with reality. Even the IFS is in on this now:

Carl Emmerson, deputy director of the IFS, said: “An ‘end to austerity’ – as defined by no further net tax rises, benefit cuts or cuts to spending on public services – would require a very sharp change of direction. 

The point being that that's not the entire budget. A notable lack there is the interest bill for the public debt. Something which has risen rather a lot in recent years and which, as interest rates rise again is going to become ever more important. We don't in fact predict that this will be true but it is certainly possible that said interest will become an expense to rival that of the NHS (debt of 90% of GDP, interest rates up to say 5%? Could happen). And we do tend to think that when we talk about a budget then we should be talking about a budget, not just the nice stuff that people like, the spending upon themselves.

We've also seen a rather more economist's definition of austerity, whatever level of spending is below what would ensure full employment. That meaning that anything less than near infinite spending in 2008/9 being austere given the depth of that recession.

The answer to the basic question is that by our measure, total spending, there has been no austerity. You can indeed cook up measures by which there has been some. But the rest of us don't have to agree with the recipe you've used to do that cooking.

Are Bank Capital Requirements Really Ten Times Higher than Before the Crisis?

A major theme in Bank of England speeches over the last three years has been the ‘Ten Times’ story: bank capital requirements are now 10 times higher than what they were at or before the time of the GFC. Here are some examples:

  • “Capital requirements for banks are much higher … In all, new capital requirements are at least seven times the pre-crisis standards for most banks. For globally systemic banks, they are more than ten times.” (Mark Carney, 2014) [1]
  •  “ … the capital requirements of our largest banks are now ten times higher than before the crisis.” (Mark Carney, 2015) [2]
  • “Common equity requirements are seven times the pre-crisis standard for most banks. For global systemically important banks (G-SIBs), they are more than ten times higher.” (Mark Carney, 2016) [3]
  • The largest banks are required to have as much as ten times more of the highest quality capital than before the crisis … (Mark Carney, 2017, his emphasis) [4]

This latter claim is particularly significant because Governor Carney is referring to the largest banks in the world and was writing in his capacity as chairman of the Financial Stability Board (i.e., as the world’s most senior financial regulator) to the leaders of the G20 countries. He could hardly have chosen a more conspicuous forum in which to make his point.

At first sight, these claims are very reassuring. After all, if bank capital requirements are now ten times greater than they were before the GFC, that must mean that our banks are now much more resilient, right?


Let’s consider the evidence. 

The evidence for Governor Carney’s claims would appear to be the capital requirements in the following Table (Table B.2) from the Bank’s July 2016 Financial Stability Report:

Notes to Table B.2:

(a)   Expressed as a proportion of risk-weighted assets. An additional 1.5% of risk-weighted assets must be held in at AT1 [Additional Tier 1 capital] as part of the Basel III Pillar 1 requirement. UK banks are also subject to Pillar 2A requirements.

(b)   See Caruana, J. (2012) ‘Building a resilient financial system’, available at

(c)   in a standard environment.

This Table indicates that the minimum Basel II core Tier 1 (CT1) capital requirement was 1 percent using Basel III definitions. The lines below show the additional requirements for the ratio of Common Equity Tier 1 (CET1) capital to Risk-Weighted Assets (RWAs), which sum to 9 to 11.5 percent depending on the settings for the systemic and countercyclical capital buffers. The systemic buffer is likely to have an impact of no more than 0.5 percent of RWAs, however.[5] As for the countercyclical buffer, the Bank of England announced on June 27th that this buffer would be raised from 0 percent to 0.5 percent. Therefore the actual value of the ‘Basel III CET1 minimum with buffers’ term at the bottom of the table should be no more than 8 percent, but let’s call it 8 percent to be on the generous side.

One can then say that CET1 Pillar 1 capital requirements involving RWAs are currently 8 times their Basel II counterparts. One can also say that the system envisages the potential for CET1 ratio capital requirements to be 11.5 times their Basel II counterparts – and even higher if one takes account of higher systemic buffers or a higher countercyclical capital buffer or the Pillar 2A requirements mentioned in Note (a) of the Table. 

At first sight, such an increase in capital requirements might appear impressive. But consider the starting base. Under Basel II, RWAs could be a hundred times bank capital. When calculating its supplementary leverage ratio buffers the Bank uses a working assumption that the ratio of RWAs to total assets is 35 percent,[6] and 35 percent is also a rough approximation of the empirical ratio of RWAs to total assets across the UK banking system. Applying this ratio, total assets might have been 100 ÷ 35% = 285.7 times capital: banks could be leveraged by a factor not far short of 300 under the old rules. Given that UK bank CET1 ratio capital requirements are currently 8 times what they were before the crisis, current requirements would still allow banks to be leveraged by a factor of 285.7 ÷ 8 = 35.7. This is a high level of leverage and high leverage was a major contributor to the severity of the crisis. 

And I have not taken account of how UK banks could increase their leverage further by switching into assets with lower risk weights or by moving positions from their banking books to their trading books.

The bottom line is that a large percentage increase in capital requirements does not represent a large absolute increase in capital requirements if the base is low to start with.

And why was the base so low? Because Basel II imposed extremely low minimum capital requirements. Correctly interpreted, Governor Carney’s ’10 times’ narrative (or to use the more accurate figure, an ‘8 times’ narrative) does not imply that banks now face high capital requirements; it is, instead, a damning indictment of the inadequacy of Basel II.

One can also look at this issue another way. The capital ratios that matter are not those based on the highly unreliable RWA measure: the ratios that matter are the leverage ratios. Basel II had no minimum required leverage ratio and Basel III introduced a minimum required leverage ratio of 3 percent. But this 3 percent minimum required leverage ratio is specified with Tier 1 capital as the numerator[7] and the leverage exposure as denominator. When one converts this leverage ratio into the ratio of CET1 to total assets using Basel rules and recent data for UK banks, the minimum ratio of CET1 to total assets is about 2.4 percent, allowing for a leverage factor of over 40.

Therefore, one can say that when it comes to the leverage ratio, the Basel III requirements are not 8 times or 10 times or even 20 times what they were: they are infinitely greater than what they were. Even so, they are still too low.

It is not for nothing that Martin Wolf has described Basel III as the mouse that did not roar.[8]

Slinging around multiples of capital ratios is great fun, but there is a serious side too. The question one must ask is why does the Bank choose to emphasise this 10 times narrative to make their point that UK banks are now strong again, when the underlying facts on the ground – the empirical leverage ratios (see here or here) – do not support that narrative.

To illustrate, consider the following chart (Chart B.2) from the BoE’s November 2016 Financial Stability Report

Major UK Banks’ Leverage Ratios

Sources: PRA regulatory returns, published accounts and Bank calculations.

Notes to Chart B.2:

(a)   Prior to 2012, data are based on the simple leverage ratio defined as the ratio of shareholders’ claims to total assets based on banks’ published accounts (note a discontinuity due to introduction of IFRS accounting standards in 2005, which tends to reduce leverage ratios thereafter).

(b)   Weighted by total exposures.

(c)  The Basel III leverage ratio corresponds to aggregate peer group Tier 1 capital over aggregate leverage ratio exposure. Up to 2013, Tier 1 capital includes grandfathered capital instruments and the exposure measure is based on the Basel 2010 definition. From 2014 H1, Tier 1 capital excludes grandfathered capital instruments and the exposure measure is based on the Basel 2014 definition.

This chart shows that in terms of actual capital-to-asset ratios, we are roughly back to 2002 levels and about 1.5 (not ten) times higher than 2006-7, the eve of the crisis. And these are book-value figures. In terms of market values, the Bank of England’s own data suggest that UK banks’ capital ratios are well below what they on the eve of the crisis.[9]

Ten times capital requirements or no, UK banks are still far from resilient. One can only hope that they will not have to go through another major stress any time soon. 


End Notes

* I thank Anat Admati, Tim Bush, Jim Dorn, James Ferguson, Gordon Kerr and Sir John Vickers for helpful discussions on this topic.

[1] M. Carney, “The future of financial reform,” 17 November 2014, p. 4.

[2] M. Carney, public statement made on the morning of June 24th 2015 shortly after the result of the Brexit vote was announced.

[3] M. Carney, “Redeeming an unforgiving world,” 26 February 2016, p. 8.

[4] M. Carney, Letter to G20 leaders, 3 July 2017, p. 1.  As an aside, I have a bone to pick with Governor Carney’s statement that banks “have” capital. To say that banks “have” capital (or that they “hold” or “hoard” capital, etc., other common errors of the same nature) is to suggest that capital is an asset to a bank and thereby subscribe to the ‘capital is a rainy day fund’ fallacy that has been debunked by Admati and Hellwig. Capital is a source of finance to a bank, not an asset. Banks do not “have” capital; they issue it. Dr. Carney confuses what banks invest in with how they finance themselves, and it is important to get these things right.

[5] The situation is however quite complicated and I won’t attempt to summarise it here. Instead, I would refer the reader to J. Vickers “The Systemic Risk Buffer fro UK Banks: A Response to the Bank of England’s Consultation Paper,” Journal of Financial Regulation, Volume 2, Number 2, pp. 264-282.

[6] See Bank of England (2015) “Supplement to the December 2015 Financial Stability Report: The framework of capital requirements for UK banks,” box 1.

[7] This Tier 1 capital measure is Basel III specific and is not to be confused with core Tier 1 or Tier 1 as defined under Basel II!

[8] M. Wolf, “Basel: the mouse that did not roar,” Financial Times, 10 September 2010.

[9] To illustrate, the market-value Simple Leverage Ratio, the ratio of Shareholder Equity to Total Assets, fell from 8.0 percent going into 2006 to 5.28 percent in November 2015, representing a decrease of 34 percent. See K. Dowd, "The Fiction of the 'Great Capital Rebuild'," Adam Smith Institute blog, July 6th 2017.

Nothing like making political capital out of a tragedy, eh?

Dawn Foster tells us that Grenfell Tower is all because of, well, umm, markets? Fatcher? Neoliberalism perhaps

After the fire, as details emerged about the intricacies of how the blaze progressed, the focus zoned in on such things as cladding and the provision of sprinklers. But survivors are clear that the inferno was not just a freak accident but the result of decades of neglect and poor policymaking; an indictment of how Britain houses its poorest people.

Across the UK, many others are suffering similar effects of the housing crisis. It has never been a crisis purely of supply and demand, but of shifts in legal tenure, the erosion of housing rights, the decimation of legal aid, the mass sell-off of social housing, and a growing callousness in attitudes towards vulnerable people.

Something did quite obviously go wrong and we'd love to find out what it was. Which we will and then we'll know and it won't be done again. But this isn't in the slightest about the general structure of Britain's housing market:

For Grenfell Tower survivors, empathy was plentiful at first, as the donations and flood of volunteers rushing to west London showed. Then came the chiding calls not to politicise the tragedy, as survivors themselves stated publicly that their ordeal was political, and the snide backlash when the City of London corporation announced they had set aside 68 flats in a Kensington development for survivors.

Let's remind ourselves of the background here. This social housing, this social rather than market asset that it is claimed we should behaving. Grenfell was exactly that social asset, it was social housing. At non-market prices. When it did burn down then equal if not better housing was found a mile or so away, right in the centre of one of the world's most expensive cities. People will be on the same tenure terms as they were.

Again, there has undoubtedly been a failure, that building going up like a torch has shown that. But this isn't a failure of Britain's housing market in the slightest. Nor has the coping with it been badly done. It's actually worth remarking upon quite how well that provision of non-market housing has worked.

There is another point we should make from the evidence Ms. Foster presents us with:

The 80 people who died in the disaster and those who escaped the fire are at the extreme of the spectrum, but currently, there are almost 120,000children homeless or living in temporary accommodation in England.

There are 120,000 children classified as homeless, not 120,000 children actually homeless. This distinction matters. For we have a system to find homes for those who don't have them. Entry into the system requires that one be declared to be homeless first. Thus the classification of homelessness is a necessary precondition of the system which deals with the problem swinging into action. This is of course Worstall's Fallacy all over again, looking at the size of the original problem, not the effectiveness (or not) of the system we have to deal with it.

The residual problem is those sleeping rough. Of whom there are some 4,000 nationally (no, not children, total) at any one time. And, as Ms. Foster tells us, this is a rather different problem:

The footballer turned property developer Gary Neville admitted to the Guardian this week how difficult it was to provide the kind of help rough sleepers need – 40 of the group Neville and Ryan Giggs allowed to stay in their vacant Manchester hotel after it was squatted were rehoused, with 20 ending up back on the streets.

As a little prompting will elicit from anyone involved with aiding those rough sleepers. The prevalence of addictions and mental health problems (variously estimated at up to 75% of this population) means that the difficulty is in keeping them in accommodation, not finding it for them.

There are undoubtedly problems in Britain's housing markets. The Town and Country Planning Act comes to mind. Fire regulations equally. But we have a social housing system, one that really works rather well. We can see that from the residual of the problem after it swings into action. We have many classified as homeless, we have many at risk of it, and absent drug, alcohol and mental health issues we have just about no one who is actually homeless.

That's actually pretty good for government work.

Universities need skin in the game

Derided at the time (and by many today) the Conservative-Liberal Democrat decision to treble tuition fees is signicantly underrated. As I pointed out, Jeremy Corbyn was wrong to state that £9k fees had put working class applicants off a University education. In fact, high tuition England has done better than tuition free Scotland at getting disadvantaged students to apply.

But the system is far from optimal.

Applicants often make poor degree choices, let down by poor quality information on employability. Indeed, there are 23 universities where graduates actually earn less than non-grads.

But that’s not the only problem. Income contingent loans mean that higher earners pay back more than lower earners. This is what makes this system ‘progressive’. Student Loans are underwritten by the taxpayer and in four fifths of cases the taxpayer ends up writing off some debt. The IFS estimates that around 43% of the total debt goes unpaid.

And there’s a massive difference between the winners and losers. If you end up in the bottom 20% of graduate earners it’s likely that on average you’ll never make an annual repayment of more than £500. Meanwhile the top 20% of graduate earners can expect pay as much as £5,000 in a single year in their 40s.

This creates a dangerous moral hazard. By reducing the risk to graduates whose degree choices fail to pay off and shrinking the reward to the graduates who chose wisely, income contingent loans create an implicit subsidy to courses that offer lower economic returns.

This is not just a problem with the student’s incentives. From the perspective of a university £9k a year from a Media Studies student who’ll graduate to earn less than £21k a year is worth just as much as £9k a year from a Physics and Engineering graduate who’ll go on to found the next Tesla. In fact, because science courses on average cost more to run than humanities courses, the Media Studies student is actually a more profitable option for the university.

IFS data shows that while average university funding is up by 25%, there’s a big divergence in where it’s going. The highest cost courses have seen only a modest 6% increase, while lower cost courses have seen a 47% increase. The Government can and does mitigate this effect with funding grants, but those grants have shrunk across the board since tuition fees were trebled.

As student numbers are no longer capped, the unintended consequence of a progressive repayment system is that universities are incentivised to push low-cost, low-return arts courses at the expense of high-cost, high-return STEM courses.

Beyond alumni donations and looking good in league tables, universities fail to capture the upside of investing in the employability of their students. This needs to change.

One suggestion from Ryan Bourne resurrects a relatively untried proposal from Milton Friedman. Rather than students taking out loans, universities would buy shares in a student’s future income. If you’re a bright student with 3 A Stars at A-Level then Cambridge might offer to provide a full education in return for a 3% stake in your future earnings. Cambridge could then tap into that income stream right away by selling it on the open market.

In some ways, this model already exists. App Academy, a 12-week intensive coding bootcamp, doesn’t charge fees but instead requires that students pay App Academy 22% of their salary for the 12 months after graduating (provided they get a job as a software engineer). This gives them a strong incentive to link up with employers, assist in the job search and most importantly, teach coding well.

Moving to the App Academy model aligns the incentives of graduates, taxpayers, and universities. It would disincentivise institutions from admitting students to low-return courses and incentivise investment in more expensive STEM courses provided they paid off for graduates.

Data like the IFS’s that simply compares large blocs of people doesn’t take account of their pre-existing skills and abilities: perhaps they would have done better if they’d never gone to university at all. This system would give students an incentive to only do degrees that enhance their earning power—since they will have to tithe to their institution.

Unfortunately, this policy may be too radical; too big of a change overnight. But discussing and studying the ideal is useful because it can reveal the direction in which policy should move in the short-term.

Here’s a more modest step.

Students should make payments direct to universities not the Student Loan Company.

Under the status quo student loan repayments end up going to the same place, regardless of whether a student went to Oxford or London Metropolitan. In effect it means that once you graduate universities have no stake in you succeeding or failing.

We should change that. Rather than each graduate making repayments to the Student Loan Company, they should pay the university directly under the same repayment terms. Universities should then be allowed to sell off that future income stream for cash today.

This would make a difference for two reasons.

First, it’d create a strong market incentive to accurately estimate graduate earnings. In order to maximise revenue from selling off the rights to future repayments, the university will need to provide accurate, evidence-based estimates of graduate earnings. Similarly, financial institutions will rigorously test each university's numbers, ensuring that they’re accurate.

Second, it would give universities skin in the game: incentivising them to invest in employability, shut down courses that don’t deliver for students, and shift resources to high-payoff STEM courses.

The 'loan' could still be written off after 30 years, but it would be the university not the government writing it off. This would amount to a cut in funding for universities, but importantly it would be a cut that hit the least useful courses and institutions hardest and incentivised expanding the courses that pay off the most. Cutting university funding would be no bad thing either. Subsidised tuition encourages universities to over-invest in new infrastructure and prevents them from making the same efficiency savings that other similar sized organisations might make.

For instance, there’s no incentive to introduce shorter courses; university terms are still based on pre-industrial revolution agricultural seasons.

There is a risk that the courses cut will be disproportionately ones that attract pupils from disadvantaged backgrounds. But we shouldn’t be focused on inputs (i.e. number of disadvantaged students attending university) but outputs (i.e. will they get a good graduate job). The courses that get cut will look good if you simply measure inputs but bad if you measure outputs. There's nothing progressive about saddling a young person with massive debts to pay for a university education that does little to nothing to improve their life chances.

By shifting the burden of funding university tuition fully onto universities and students, it will free up additional public money that can be used to address educational inequalities earlier on - an approach that could be more fruitful.

If we want universities to serve their students better, they need to have skin in the game.

The Coming Moral Panic About Sex Robots

Last week, The Salvation Army released a statement expressing concern that sex robots could increase demand for sex work and sex trafficking. This particular moral panic seems a little premature; according to Nature, “just four companies, all located in the United States, currently produce [extremely basic] sex robots.” But this hasn’t stopped some social conservatives and feminists uniting in opposition to the potential spread of this emerging technology.

Unsurprisingly, rigorous academic studies into the effects of sex robots are extremely hard to come by. But the battle lines have already been drawn—anyone familiar with other debates relating to the sex industry (e.g. sex work and pornography) knows that this research area is plagued by motivated reasoning, blind speculation, and emotive anecdotes. Sadly, I do not think that the coming debate on sex robots will be any different.

British opposition to sex robots is led by the Campaign Against Sex Robots, spearheaded by robotics researcher Dr. Kathleen Richardson. In the position paper that launched the campaign, Dr. Richardson wrote that:

“The arguments that sex robots will provide artificial sexual substitutes and reduce the purchase of sex by buyers is not borne out by evidence. There are numerous sexual artificial substitutes already available, RealDolls, vibrators, blow-up dolls etc., If an artificial substitute reduced the need to buy sex, there would be a reduction in prostitution but no such correlation is found.”

But if there is no correlation between the availability of artificial sex substitutes and the amount of sex purchased, then this also rules out the possibility that sex robots will increase demand for the purchase of sex! For the moment, the arguments on both sides are speculative. Richardson states that “new technology supports and contributes to the expansion of the sex industry,” citing the growth of the sex industry spurred by the expansion of internet. To me, this seems like a very weak argument; there is an obvious, meaningful difference between the internet’s effects on human sexual commerce and sex robots’ potential effects on human sexual commerce.

My prediction is that, like sexual violence and pornography, the substitution effect will dominate. I also predict that this substitution effect will be larger for sex buyers who aren’t as interested in the mutuality aspect of commercial sex: arguably a more problematic group of sex buyers.

Let’s say I’m wrong, and that sex robots turn out to be a complement to buying sex. Would an increase in demand for sex work brought about by sex robots necessarily be a net harm to society? The whole ‘End Demand’ approach to sex work is fundamentally flawed, and any potential harms must also be weighed against potential benefits such as using sex robots to alleviate loneliness and assist in sexual therapy.

It’s also worth noting that whilst men are almost certainly going to be the primary market for sex robots, they aren’t the only group involved. Dr. Richardson briefly highlights this in her position paper:

“But the development of sex robots is not confined to adult females, adult males are also a potential market for homosexual males.”

Women of all sexualities are also likely to comprise a non-trivial proportion of sex robot owners. In the few surveys conducted into attitudes towards sex robots, women “answered positively about half as often” as men. The idea that some women may purchase sex robots as they become more widely available is not that farfetched.

In the coming years, the debate over the legal and societal approaches we should take towards sex robots will become more prominent. That conversation must include voices that emphasize their potential positive impacts and call out evidence-free scaremongering.