Elections do not give you the wisdom of the crowd

In his column on Monday, Matt Ridley argues that a majority of voters tend to come to the right answer surprisingly often, and even beat the experts.  The mileage you give that view will likely vary depending on what you think of Trump and Brexit. But what's interesting is why he thinks this. Ridley argues that elections benefit from something known as the wisdom of the crowd, he writes:

‘In these democratic days, any investigation into the trustworthiness and peculiarities of popular judgments is of interest.' So begins an article entitled Vox Populi, which is not about Donald Trump but was published in 1907 by Francis Galton, a pioneer of statistics, by then 85 years old. He had analysed the results of a sweepstake competition held at the West of England Fat Stock and Poultry Exhibition in Plymouth.
An ox was on display. Visitors could buy a postcard for sixpence and write their guess as to the weight of the ox, once slaughtered and dressed. Of 800 cards filled out, Galton rejected 13 as illegible and averaged the rest. The arithmetic mean of the 787 guesses came to 1,197lb. The true dressed weight of the ox was — yes — 1,197lb (Galton reported slightly different results, but recent reanalysis by Kenneth Wallis of Warwick University finds the match was exact).
The message is that a crowd is at least as wise as any expert (only one guess was spot on). In a large group of people, ignorance in one direction cancels out ignorance in another."

It's a delightful story and arguably one of Galton's most interesting discoveries (arguable because as Scott Alexander points out Galton gave us "several statistical tools including correlation and standard deviation, the use of the survey in data collecting, the term “eugenics”, the entire science of meteorology, hearing tests, the first study on the power of prayer (he prayed over random fields to see if the crops there grew higher; they didn’t), fingerprinting, the scientific investigation of synaesthesia, and a horrible warning about how not to do facial hair"). 

Ridley points to another area where the crowd beats individual experts – markets. They tend to outperform individual forecasts with remarkable–but not perfect– accuracy. It's why we at the Adam Smith Institute argue that monetary policy shouldn't be determined by a panel of wise men and women, but by a Nominal GDP futures market.

But there's a problem.  Contrary to Galton's example, in elections a voter's ignorance in one direction doesn't cancel out ignorance in another. Suppose that there was some sort of systematically incorrect belief that Oxen were 50lbs heavier than they actually are; maybe the guessers were primed with stories about heavy oxen. If that was the case then the mean guess would no longer give you right answer.

I think that this is probably the case with voting. Bryan Caplan in his excellent book The Myth of The Rational Voter lists four biases (it might be better to think of them as strong errors) that systematically lead voters to make a sub-optimal choice.

First, there's the make-work bias. People tend to confuse economic growth and job creation, but most of the real benefits of economic progress come from saving labour. It leads voters to favour government programs that make-work to the most economically beneficial ones.

Second, you've got the anti-foreign bias. People significantly underestimate the benefits of free trade and migration. Trade is often phrased as zero-sum with winners and losers (note Donald Trump's insistence that China's killing America by providing US Businesses with cheap steel and US consumers with cheap clothes). The reality is that trade deals are win-win, some winners are bigger than others but both sides are still winners.

Third, voters are overly pessimistic about economic conditions. Despite the tireless and highly commendable work people like Matt Ridley, voters systematically underrate the state of the economy in both the short and long term.

Fourth,  voters are systematically biased against market mechanisms. They tend to be overly suspicious of the profit motive. Caplan cites oil prices as an example. 74% of Americans blame them on greedy oil companies, just 11% of economists agree.

Ridley's right that a council of experts simply isn't as good as the wisdom of the crowd, but it's markets not elections that we'll get it from.

There's economics and then there's not economics really

In a tale from Oop North we're told that the destruction of manufacturing employment and the coal industry now costs the Treasury some £30 billion a year in extra welfare benefits and lower tax revenues. This comes as some of the resultant unemployment is disguised as incapacity benefits and then the various top ups to incomes resulting from those lower paid service jobs. The Treasury here meaning of course our own pockets as we must pay tax to support those who used to support themselves through making things and paying tax as they did so.

This is not a convincing analysis of recent events. No, not even when The Guardian reports on it:

The enduring impact of closing factories and shutting coalmines in the 1980s has been revealed in new research showing that the drain on the exchequer from former industrial areas is responsible for up to half the government’s £55bn budget deficit.

In the first comprehensive analysis of the cost to the state of the de-industrialisation that began three decades ago, Sheffield Hallam University said the annual bill was at least £20bn and was perhaps as high as £30bn.

The economic cost of something, whatever it is, is the net cost of something, not the gross. With climate change there's damage to the future from emissions – and benefits in the present from making those emissions. It's the net cost that Lord Stern spent all that time calculating. The net cost of inequality is the difference between the impetus to strive it provides minus the damage, if any, inequality itself does.

So it is with manufacturing or coal mining. What were the costs of supporting these activities so that people worked and paid tax? What is the net cost to the Treasury, or our wallets, not what is the gross?

At this distance we're not going to be accurate about such costs of supporting those industries. But as a guide:

Taxpayers were subsidising the mining industry to the tune of £1.3 billion annually. This figure doesn’t include the vast cost to taxpayer-funded industries such as steel and electricity which were obliged to buy British coal.

Using the ever indispensable Measuring Worth we find that that £1.3 billion is some £5 billion a year now. And that is, again we should note, just the direct subsidy to coal mines, not the amount also picked from our pockets by higher input costs to British industry and our electricity bills. We can all have the most lovely arguments about how much further we want to go in adding the costs of supporting that industry – as someone around at the time I would want a hefty addition for people having to drive Morris Marinas. But it is indisputable that the net cost was very much lower than the gross cost being offered to us in this report.

All of which is a pity really for as the report itself, from the Centre for Regional Economic and Social Research, says:

Second, there is an urgent need to discuss the role of evidence in the formulation of public policy. Policy making rooted in evidence and analysis is out of favour. Political debate appears more interested in appealing to emotion, speculation and imagery. Facts are seen as irrelevant and experts as dour pessimists. This rejection of evidence in favour of supposed ‘common sense’ thinking in practice risks distancing policy from the lived realities of the people and places it should be serving. The result is policy that misunderstands what is going on, does little to make things better and can often make them worse.

Yes, that would be both interesting and useful, wouldn't it? An analysis of the costs and benefits of political plans, of economic policy choices. Perhaps we should look to that generously supported academia of ours to produce it.

But perhaps not at Sheffield Hallam as whatever it is that they're doing up there it doesn't appear to be economics. Nor hugely useful if we are to be candid.


Deliveroo and why the gig economy works best

In the wake of the recent tribunal which declared that Uber drivers should be classed as employees with rights to the minimum wage, holiday pay and so forth, we read today that something similar is happening with Deliveroo, as riders seek to unionise and gain similar workers' rights as Uber.

Deliveroo is a company that provides a delivery service on behalf of thousands of restaurants across the country, and it classes its riders as self-employed "independent contractors", on a pay rate of £3.75 per delivery.

It is true that under such conditions they lack some of the regular workers' rights many of us have, such as paid holiday and the right to the minimum wage, but it's incredibly short-sighted to assume that that means there is a problem needing fixing - after all, by equal measure, Deliveroo contractors also enjoy some working benefits most of us do not; namely, the flexibility to work the hours that best suit them, and as little and as often as they wish.

Rather like how the decision made against Uber was so drastically wrong, the same will apply here, because there is an abject failure to understand that in a dynamical and diverse market, a 'one size fits all' imposition frequently harms little niche elements that are doing perfectly well as they are.

The ability of Uber and Deliveroo workers to be classified as flexible, autonomous independent contractors ostensibly running their own businesses as available drivers and riders will be, for many, a much more beneficial set-up (ditto numerous others on zero hour contracts) than the myopic imposition of terms and conditions that would no longer favour their flexible working patterns.

When left alone, firms like Uber and Deliveroo are beneficial to the economy (for employer and consumer) precisely because they involve conditions that those who favour them will take advantage of, and those that do not, won't. Even under the new terms of £4.25 per ‘drop’, most Deliveroo cyclists are earning significantly above the minimum wage as the average worker makes two deliveries per hour, equating to £8.50/hour.

It's hardly rocket science; people who want to earn the minimum wage and have holiday pay and less flexible working arrangement will not be Uber drivers or Deliveroo riders. On the other hand, people who want to be self-employed contractors under the flexible conditions offered by Uber and Deliveroo will. That is the very nature of a free market.

Not only is this far from rocket science, there is even a quote in the article from one of the riders that with closer examination would provide him with the answer he ought to be looking for. The Deliveroo rider laments that "We don't get an hourly fee, so that means at times when there aren't that many deliveries and it is not that busy, we can be waiting for up to an hour for a delivery without getting paid a penny." - which really ought to give the game away.

If only this young man could see that the problem with his campaign is heavily hinted at with the words he utters - for if there are times when it is not that busy, and riders can be waiting for up to an hour for a delivery without getting paid a penny, it's not exactly going to help out the situation by forcing Deliveroo to shell out a state-mandated £7.20 an hour to all its riders, is it? All it will do is skew the market value of the Deliveroo rider operations and harm all the people who benefit from riding for Deliveroo under the present conditions. 

It must be very annoying being George Monbiot

It's not really possible to determine quite how many bees there are bouncing around George's bonnet but there are three that do stand out.

One is a noted antipathy to private property and markets as a method of organising the world.

Another is the spitting rage at the manner in which property ownership is so concentrated into few hands in this country. 

A third is the idea that we should stop farming and maintaining many of the formerly wild places in this country and "rewild" them.

We have a certain sympathy with that third idea and would most certainly love to see the lynx back, if perhaps not the bear or the wolf but we're open to that idea too.

But to what must be annoying about holding all three views:

Not since the Viking era has a Scandavian taken so much British land - but Denmark's richest man is now not far away from owning more private property in the UK than anyone else.

Billionaire Anders Holch Povlsen  has continued a personal crusade of snapping up swathes of the Scottish Highlands by purchasing the 18,000-acre Eriboll Estate.

According to The Times, his mission stems from a dream childhood holiday, and he now owns 218,364 acres of private land on our shores, only just bettered by the Duke Of Buccleuch's 236,000 acres.

But, as opposed to many modern developers who aim to make huge profits with development, Povlsen is seeking to return his land to nature with a huge rewilding process that will stretch over 200 years.

It must be just excruciating to find that it is precisely the concentration of property into the hands of one owner, that concentration only possible through the institution of private property and markets, which allows anyone to overcome the coordination problems of rewilding.

For us, well, it's his property he can do what he likes with it. That's what private property means. But it must just grate for George and others that the desired goal is being achieved through not Goodthink means.

Markets still don't like racism

As I've written before, you'd expect competitive markets to drive out racism. If firms have "tastes" about which workers they will hire, or they value labour from some groups at less (or more) than its true product, they will be driven out of business. This is the theory—but it's also the evidence. And yet more evidence has added to that we already had.

The first paper, an NBER working paper by Nathaniel Hilger, asks why Asians have high income despite a history of intense racism directed towards them. He finds:

Asian Americans are the only non-white US racial group to experience long-term, institutional discrimination yet today exhibit high income. I reexamine this puzzle. I focus on California, where most Asians settled historically. Asians achieved extraordinary upward mobility relative to both blacks and whites for every cohort born in California since 1920. This mobility stemmed primarily from gains in earnings conditional on education, rather than unusual educational attainment. Historical test score data suggest that low initial earnings for Asians—unlike blacks—primarily reflected prejudice rather than skills. Asian history is consistent with the view that racial earnings gaps driven by contemporary prejudice do not persist in sufficiently competitive labor markets.

Asian Americans used to be paid less than their test scores would suggest, but this evened out over time as competitive markets drove out prejudice. Another paper, by Devah Pager, looks at the same question by focusing directly on bigotry and firm failure:

Economic theory has long maintained that employers pay a price for engaging in racial discrimination. According to Gary Becker’s seminal work on this topic and the rich literature that followed, racial preferences unrelated to productivity are costly and, in a competitive market, should drive discriminatory employers out of business. Though a dominant theoretical proposition in the field of economics, this argument has never before been subjected to direct empirical scrutiny. This research pairs an experimental audit study of racial discrimination in employment with an employer database capturing information on establishment survival, examining the relationship between observed discrimination and firm longevity. Results suggest that employers who engage in hiring discrimination are less likely to remain in business six years later.

Turns out markets are fairer than you might think!

HyperDramatisation: How Adam Curtis was consumed by a fictional reality

The thesis of Adam Curtis’ HyperNormalisation is that over the last 40 years technocratic risk managers and utopian fantasists have fed us a simplified fictionalised reality. The notion that suicide bombers, developments in AI, austerity, “cyberspace utopianism”, Prozac, the efficient market hypothesis, and your newsfeed share an origin story is certainly an audacious proposition.

As such, Curtis has to treat very ordinary political phenomena as shocking, revolutionary developments. Did you know, for instance, that governments lie? And international alliances change? And Russia disseminates propaganda? A lot of events are reduced to caricature. The Syrian civil war can apparently be traced directly and unambiguously to Henry Kissinger’s realpolitik approach to the Middle East peace process. Other important contemporary issues, such as Egypt’s shifting agenda following Gamal Adbel Nasser’s death in 1970 and the precarious position of Syria’s ruling (Shia) Alawite minority, are not mentioned.

Curtis seeks to frame the modern world as peculiarly defined by disillusion, deception, cynicism, manipulation, and injustice. However, there’s no reason to think any past period was somehow different. Curtis suggests that a significant transition happened in the 1970s, focusing on New York City’s fiscal crisis and its subsequent ‘takeover’ (in reality, budget management) by the Financial Control Board.

Curtis’ characterisation of the ideology purportedly behind this (he never uses the word ‘neo-liberalism’) as being premised on belief in stable systems and the infallibility of predictive modelling is a major distortion. If anything, the best (and not infrequently made) argument for free markets is that economic planning is impossible and price signals are necessary precisely because the future is unpredictable and the economy is never in stasis.

Curtis also bemoans a purported retreat of radicals away from politics and into art and irony. He fails to make a convincing case that, to the extent that this happened, it’s actually negative. Curtis quotes Patti Smith criticising large collective movements as stifling and perversely bureaucratic. Nothing suggests she was wrong, and it is odd to think she contributed less to humanity by making music than she would have done if she’d spent her time reading Trotsky. Occupy and the Arab Spring are portrayed as failing because they organised spontaneously and eschewed hierarchy. Revealingly, Curtis refrains from actually endorsing explicitly hierarchical social movements.

The implication that progress can only be achieved through government and collective movements is pernicious. Progress happens incrementally and experimentally in a deregulated environment. For all the dismissal of ‘techno-utopianism’, the Internet has provided enormous benefits for millions of people. If this comes at a price of more sophisticated marketing and people indulging in narcissism, trivia, and echo chambers then it’s still a good deal.

If Curtis is describing something novel, it’s only greater access to information. Reagan did not invent the cultural myth, Blair’s rehabilitation of Gadaffi was not the first clumsy sham marriage of realism and idealism, and the Iraq War was not the first foreign policy disaster founded on lies or mistakes. The only difference (perhaps) is that people have become more aware.

Parts of HyperNormalisation are illuminating, such as the discussion of Trump’s imperviousness to fact checking, the explanation of Vladislav Surkov’s propaganda techniques, and the revelation that Bashar al-Assad had a brother called Basil. However, ironically, by including and connecting everything that’s happened since 1975, Curtis ultimately only manages to portray a simplified, conspiratorial fiction himself.

Yes, we do know what to do it's just being allowed to do it that is the problem

Quite how schools should be run and by whom is one of the great battles of our times. The reason this is a battle is because the government pays for them thus politics defines how they should be run and by whom. Which brings us to the Academy movement, or as it is in the US, the charter school movement.

At which point we have the New York Times. Yes, they're inherently liberal in that American sense, statist even, but if even they are getting the message perhaps there is hope yet:

The briefest summary is this: Many charter schools fail to live up to their promise, but one type has repeatedly shown impressive results.

Hannah Larkin, the principal at Match, refers to such schools as “high expectations, high support” schools. They devote more of their resources to classroom teaching and less to almost everything else. They keep students in class for more hours. They set high standards for students and try to instill confidence in them. They focus on giving teachers feedback about their craft and helping them get better.

Schools that actually try to teach pupils, which monitor the ability of teachers to teach pupils, work well. Our word, fancy that.

Finally, no matter how successful charters may be, they undeniably make life uncomfortable for some people at traditional schools.

Quite so. But what are we running schools for, the education of the children of the comfort of those doing the educating? 

Let us rephrase that. We are, with the defence of traditional as it is now schooling, arguing that schooling should defend the interests of those doing the educating. But what we should be defending is the quality of the education received, not the quality of life of those doing the providing.

Or as it should be put, that free market in education methods tells us, teaches us, which method works. And our method of measurement is and should be - sod the teachers, what about the children?

It doesn't matter who wins the Presidential election, Trump or Clinton

Contrary to popular belief it does not matter who wins America's coming election. Sure, there're all sorts of noises made here and there about how it's the end of civilisation. Either way, who ever gets in, dependent upon who you listen to. But we're here to tell you this isn't true.

For America needs no more governance. There just are no more things that anyone needs to decide upon. Everything is solved. This must be true or this would not be happening:

The U.S. Food and Drug Administration (FDA) considers doing something that may affect Nutella lovers' consumption of the popular chocolate-hazelnut spread. It might change the recommended serving size on its jars. Alongside this, it wants to know how consumers eat the spread: on bread or crackers, on cupcakes, or straight from the jar?
According to CNN, the current recommended serving size is two tablespoons or 37 grams. This amount is equal to 200 calories. Ferrero, the maker of Nutella, thinks a smaller serving amount, which also means fewer calories, might make consumers more likely to buy a jar. For the past couple of years, the company has been petitioning FDA to do the changing.

Yes, really, the vast federal bureaucracy is polling the public to ask whether they use one or two tablespoon fulls as their typical serving of a hazelnut spread. No, really, it's here.

Bernard Levin once pointed out that he was just overjoyed to live in a country where the Advertising Standards people adjudicated upon the crispiness of crisps. For only a society which had solved all other problems would possibly divert resources to this sort of question.

And thus elections are of no more moment in the US. All important matters have been solved and we are now down to whether we should warn of 100 or 200 calories on a jar.

We're done, obviously, elections, government, all solved.

Why does this surprise anyone in the slightest?

Certain things in this world are just obvious. Demand rises the cheaper things become and if you make something free at the point of use then demand will be very high. Possibly even higher than you or anyone else is willing to pay for - for of course even things which are free at the point of use have to be paid for somehow.

So why this surprises anyone is unknown:

Millions of patients are putting unnecessary strain on the NHS by seeking medical help for minor complaints such as colds, insect bites and dandruff, according to a report.

Overloaded GP surgeries and A&E units are having to divert scarce resources into dealing with flu, sore backs and travel sickness, the Local Government Association said.

One of the functions of pricing is to actually ration access to such services. No, not to stop the poor from getting their due, but to stop those who don't need the services from blocking access to those who do.

Even the famously more liberal and socially democratic than we are Swedes have got this message. A visit to the equivalent of a GP costs money, real money. Not a lot, it is true, and there's a cap on the annual amount but enough to make people think about whether a sniffle best treated with a handkerchief is really worth some portion of an hour's time from a paid professional.

It works too - as anyone with even a passing knowledge of supply, demand and the role of prices in balancing them would know from the outset.

Sadly, we're ruled by those who don't know these things, aren't we? Or at least our health care is managed by those who don't.

Cannabis tampons: an answer to the period problem

A US company called Foria has just released cannabis-infused tampons made from cocoa butter, THC oil and organic hemp. Although they’re currently expensive and unapproved, they offer pain relief that many users say is more effective than any other form of PMS painkiller without giving off the psychotropic (“high”) effect.

It may sound gimmicky and silly, but it brings to light a worldwide (and somewhat taboo) problem women face that has huge social and economic consequences: periods. According to a study by Cleanmarine, seventeen million days of work are taken off annually in the UK due to PMS, with a third of women taking four or more days off every year. This makes sense, considering a NHS survey found that 90% of women experience period pain and 14% of women are frequently unable to work because of it.

The economic impact of PMS is somewhat hard to gauge, although it is clear that periods set back women to a greater or lesser extent across the world. A US study showed that women who have heavy periods are 72% less likely to be working as women with normal or lighter periods. This amounts to an annual $1692 income loss per woman with a heavy flow, and this figure is likely to be even higher with inflation taken into account.

The economic losses associated with menstruation are heightened in poorer countries. Nearly one in four Indian schoolgirls drop out of education as soon as they get their period, primarily because they are quickly married off. However, even the girls who stay on at school miss on average 60 days of school per year because of menstruation due to lack of sanitary products. The economic and social repercussions of women not going to school are huge, with one study showing that GDP goes up by 0.3% with every 1% increase in girls attending secondary school.

Although British women are not suffering the way women in places like rural Uganda are - where 11% of girl’s school days are missed due to periods – menstruation’s impact on workplace productivity cannot be ignored in any country. Taiwan, Japan and Indonesia all see period pains as such a limiting factor on women’s workplace productivity that they all offer for paid menstrual leave.

However, again, paid menstrual leave costs the economy and raises gender equality questions, as men may feel resentful about covering women’s lost work and women miss out on workplace experience that may help them further their career. Women may also be accused of cheating the system, particularly as it is hard to prove how much pain a person is actually in. When the concept of menstrual leave was debated at the Festival of Ideas in Cambridge, an Indonesian boss said that many of his female workers who took time off for PMS turned out to be taking time off to go shopping together.  

Of course everyone should be entitled to take time off when they are in too much pain to work, but effective pain relief – such as cannabis tampons - addresses the root of the problem of PMS in a way that legislative sick leave does not. As we can see with schoolgirls in India, when women are forced to take time off because of menstruation, it is harmful to both them and the economy. Supporting scientific innovation which allows women to get on with their life and work is therefore the answer.