The Living Wage is a false solution to our problems

I normally agree with the Centre for Policy Studies on economic issues, so I was surprised to see Adam Memon’s call for a mandatory ‘Living Wage’ last week. Philip Booth has already written a post criticising Memon’s original piece, but I’d like to add my perspective to Adam’s response to Philip, posted today.

To be clear, Adam prefers “tax cuts, deregulation and other supply-side measures to boost productivity”. He and the CPS have long argued for tax cuts for the poor. This is admirable, and as Adam says it deserves to be acknowledged.

My main contention is that Adam is comparing apples with oranges by using the impact of historical increases in the National Minimum Wage (NMW) to justify a future rise to the NMW to Living Wage levels. There is a lot of evidence against his position that he ignores.

Adam says that “an objective reading of the studies of the impact of the National Minimum Wage can only lead to the conclusion that it has boosted the incomes of the low paid without particularly damaging employment”. Correct. There does not seem to be much, if any, good evidence that the NMW has increased unemployment in the UK.

But this doesn’t tell us that employment would not be higher without rises to the minimum wage. Simply looking at times when we have raised the NMW, and looking at whether unemployment has risen or not, as Adam does once, is extremely crude – there are of course many other factors going on, and without an analysis that attempts to control for those factors we have no idea what the counterfactual would be. But, yes, there have been more sophisticated studies in the UK that do suggest that the NMW has not harmed employment compared to there being no NMW.

Adam says that “This is quite a big deal because it does rather make the traditional argument that the minimum wage would destroy jobs somewhat out-of-date”. But, unless we think there is something particularly unique about the UK’s labour market, the UK is not the only place we have to look at.

Internationally, most of the evidence is that increases in the minimum wage do increase unemployment at the margin. I looked at some of this last year:

Neumark and Wascher’s review of over one hundred studies found that two-thirds showed a relatively consistent indication that minimum wage increases cause increases in unemployment. Of the thirty-three strongest studies, 85 per cent showed unemployment effects. And “when researchers focus on the least-skilled groups most likely to be adversely affected by minimum wages, the evidence for disemployment effects seems especially strong”. There is evidence that suggests that minimum wages deter young workers from acquiring these skills that allow them to get better jobs in the long run.

Of course there are times when this does not happen, but most of the time it does. Most of this evidence is based on US data, and much of it compares employment rates in similar US states where one has had a minimum wage rise and the other has not.

Though UK evidence might be the most relevant evidence we have, we would need a very good reason to completely ignore the international evidence and suppose that the UK experience is all that we should look at.

I am certain that Adam agrees, because he has cited international evidence in discussions about the UK in the past. And rightly so.

Is the UK special? Maybe. But the Low Pay Commission seems to disagree, because its recommended increases have been very low compared to what Adam is proposing. Similarly, the Living Wage Foundation does not call for a mandatory Living Wage.

Distributionally, if some people are put out of work but others receive pay rises, this may well be a negative. Adam says that “There are of course some who lose out from the minimum wage but there are many more who benefit”, but concludes that “broadly speaking the minimum wage is a net positive.”

But taking all of one person’s earnings and distributing them among other people who are already in work is likely to be harmful overall, because of diminishing marginal utility. If there is an unemployment effect it may well be an upwards income redistribution from now-unemployed people to the people who hang on in their jobs.

I do think the Low Pay Commission has done a good job at keeping NMW increases quite restrained. That’s why I suspect they would balk at the idea of raising the NMW to the Living Wage level for the foreseeable future. It’s simply not convincing to compare previous rises that the Low Pay Commission has deemed safe with a future rise that it presumably deems unsafe.

Note that productivity has been very low recently, and the Low Pay Commission has barely raised the NMW as a result.

I find it extremely implausible when Adam defends his claim that the Living Wage might lead to extra productivity gains from workers. This concept is known as ‘efficiency wages’ – a well-paid worker is often a more profitable one.

But firms are profit-seeking, so wouldn’t they be doing this already? Adam addresses this by saying that “often some of these productivity gains through eg reduced absenteeism are unanticipated by firms because unsurprisingly, they don’t always have perfect information” – fine, but these firms will be the exception, not the rule. Yes, firms sometimes miss out on profit opportunities – this doesn’t mean that I or Adam or anybody else knows better.

I enjoyed Alex Tabarrok’s recent post on this, “The False Prophets of Efficiency Wages”. He points out that ‘efficiency wages’ were actually studied by economists as a way of explaining unemployment:

In the original efficiency wage literature there is no wishful thinking–no idea that we can have more of everything that we want without tradeoffs. Instead of being desirable, the efficiency wage is a problem because lower wages would reduce unemployment and be better for the economy as a whole.…

Firms routinely track turnover and productivity and they are well aware that higher wages are a possible means to reduce turnover and increase productivity although, as it turns out, not necessarily the most effective means. Indeed, the whole field of workforce science deals with retention, turnover and job satisfaction and the relationship of these to productivity and it does so with more nuance than do most economists. Thus, it’s simply not plausible that large numbers of firms on the existing margin can increase wages, profits and productivity.

To be fair, Adam suggests that his Living Wage rise would be offset by cuts in taxes for business. If these were specifically cuts to the cost of hiring workers this may actually work: cutting employer NICs for NMW workers workers might offset the extra cost of paying the worker the Living Wage. But this would just be a roundabout way of cutting the income taxes or employee NICs of those workers. The Living Wage would be doing none of the heavy lifting, and would still exclude some workers from jobs.

Adam claims that tax credits and other in-work benefits subsidise employers by letting them pay their workers less. I’ve always found this a strange claim. Why would workers’ wage demands fall just because they’re getting top-up money from elsewhere? Do lottery winners ask for lower wages? In any case, he does not provide evidence of this. The consensus from the literature I have seen is that both payroll tax cuts and wage subsidies go to the workers, without driving down wages. So there is no subsidy effect.

In light of all this, my basic view is that raising the minimum wage always risks creating unemployment, and raising it as high as Adam wants would run a very large risk of creating unemployment. I believe that low pay will be the economic problem facing my generation, as unemployment was for my parents’ and grandparents’ generations. To address it, I prefer cash transfers like the Basic Income and anything that boosts innovation, so we can improve people’s productivity and the total stock of wealth.

At best the Living Wage will act as a roundabout way of cutting taxes on workers. At worst it will put many people out of work. I admire Adam’s willingness to challenge the orthodoxy on our side, but in this case I believe that the bulk of the evidence in favour of the free market orthodoxy. The Living Wage is a siren call – a seductive but false solution to the problem of low pay. We should reject it.

The case against caring about inequality at all

Readers of this blog will probably not need convincing that inequality is not something to worry about. We’re more interested in reducing absolute poverty. If you become £100 richer, and I become £50 richer, I say that’s a good thing. But because we’ve become less equal, someone who is concerned with inequality alone would not.

But even given this, inequality might matter. Whether we think they should care about it or not, people do, and it makes no more sense to think of that as a ‘bad’ or unimportant desire than thinking a passion for expensive or high-tech watches is bad.

And because people care about it, they might act on it. If inequality makes a revolution or populist, anti-market governments more likely, as Noah Smith suggests it does, then it might reduce investment and growth as well.

Crucially, these harms from inequality come from people’s perceptions of inequality, not necessarily actual inequality. Which makes a new NBER working paper, “Misperceiving Inequality”, rather interesting (hat tip to Bryan Caplan, who quotes some of the key parts directly).

The paper shows that most people know very little about the extent and direction of income inequality in their societies, or where they fit in to the income distribution. This holds for wealth as well as income.

Screen Shot 2015-05-28 at 14.32.55

This isn’t a pedantic complaint about imprecision. One question asked people to choose which of five diagrams, above, best described where they live. Responses differed significantly between different countries (68% of Latvians chose Type A, 2% of Danes did), but in almost every country a majority got it wrong.

Globally, respondents were able to pick the “right” diagram only slightly better than randomly – 29% got it right, compared to a random baseline of 22.5%. Accuracy differed significantly between countries: 61% of Norwegians got it right, 40% of Britons did, 5% of Ukrainians did. In only five countries out of forty did more than half of respondents guess correctly. (All this uses post-tax-and-transfer data; people’s accuracy is much worse if you use pre-tax-and-transfer data.)

And respondents weren’t even close – looking at how many people were only one diagram off the right one, respondents only did one percentage point better than random (69% versus 68%). As the authors note, “with only five options to choose between, getting within one place of the correct option is not a very difficult task”.

The paper also shows that people are terrible at judging where they fall in the income distribution – 40% of British second-home owners said they were in the bottom half. 3% said they were in the top 10%.

Crucially, given worries about investment and political instability, “In countries where inequality was generally thought to be high, more people supported government redistribution. But demand for redistribution bore no relation to the actual level of inequality.”

There’s too much in the paper to cover in one blogpost, but the results are extremely clear: people’s perceptions of inequality are really, really inaccurate – that holds globally and in all but a handful of Scandinavian countries.

There are some good arguments in favour of reducing inequality based on how people perceive it – that it makes people unhappy, more left-wing, more prone to revolution, more hateful to the people around them.

But this paper shows that those perceptions are related to the realities of inequality only very slightly, if at all. Redistributive policies that reduce actual inequality are costly, and because actual inequality is barely related to perceptions of inequality they may do little to make the country more stable or market-friendly. If these are important problems, we can only solve them by making people feel less unequal – not by making them less unequal in fact. In short: even if people’s perceptions of inequality matter, the reality does not.

When a fossil fuel subsidy is not a subsidy

You may have seen an IMF report in the news last week claiming that fossil fuels are subsidised to the tune of over five trillion dollars every year. This made good headlines, but only because the IMF chose to describe untaxed externalities as ‘post-tax subsidies’. This is unusual and misleading. I wrote about why in The Daily Telegraph:

The IMF’s idea of “subsidies” to fossil fuels refers to something completely different. They have taken the indirect costs to society of using energy – air pollution, traffic congestion, climate change – and, if governments haven’t imposed special taxes on one, called it a “subsidy”. The problem is, we already have a word for these things: externalities. And there is something rather Orwellian about describing a failure to tax something as a subsidy. Here’s an example of what we’re talking about: when my neighbours play loud music at night, it makes me worse off. I’d pay, maybe, £20 for them to shut up, if it wasn’t so awkward to go to the flat downstairs, knock on their door and start negotiating prices. Economists would say that they are imposing a £20 externality on me, and that in a perfectly efficient world, my building would charge residents around that much to play music, and give it to sleep-deprived neighbours like me. But, in the absence of that charge, nobody would say that those neighbours are being subsidised by me. It’s just not what the word means. Except, apparently, to the IMF.

That isn’t to say that externalities should never be taxed, if a private solution can’t be found. But we already have high fuel taxes in most of the developed world, and in the developing world these taxes will hold back growth. Since economic development has positive externalities, it’s not obvious that the negative externalities of fossil fuels outweigh the positives. You can read the whole piece here. 

Mark Carney meets some journalists looking for a story

If you’ve read the papers today you might have seen the story about Bank of England Governor Mark Carney’s supposed intervention into the immigration debate. The Mail and the Times both covered Carney’s inflation report with this angle.

The Times’s coverage was boldest, claiming that Carney had ‘waded into the debate on immigration’ by describing ‘the present high level of net migration as “a key risk” to the economy’.

This is strong stuff, and it would indeed be big news if it were true. But I’m not sure it is.

Here’s what Carney actually said:

In recent years labour supply has expanded significantly owing to higher participation rates among older workers, a greater willingness to work longer hours and strong population growth, partly driven by higher net migration. These positive labour supply shocks have contained wage growth in the face of robust employment growth. Wages have grown by around 2% in the past year – less than half the average rate before the global financial crisis – and a key risk is that these subdued growth rates continue.

Such strong growth in labour supply is unlikely to be sustained. Going forward, growth in the UK economy’s potential will increasingly depend on productivity.

It’s hard to see Carney’s exact meaning from this, and he has already distanced himself from the Times’s and Mail’s interpretation. As BusinessInsider’s Mike Bird points out, he’s most likely talking about a compositional effect – the average changing because we’re adding more people on the lower end of the spectrum, not because any existing worker is being made worse off.

Bird quotes the Inflation Report itself:

Bank staff estimates suggest that the changing composition of employment growth — including the mix of occupations, industries, ages and job tenures — could explain around 1 percentage point of the recent weakness in average annual earnings growth. Compositional effects will only suppress wage growth for as long as such shifts continue.

Indeed. It would be a surprise if Carney had said what the Times and Mail suggest he said. The government’s Migration Advisory Committee found in 2012 that “Studies estimating the impact of migrants on UK wages have generally found little or no impact on average wages,” although, “in some studies migrants were found to increase wages at the top of the UK wage distribution and to lower wages at the bottom.”

That means that, if there is a negative impact from immigration, it reduces wages for the workers at the bottom, but not the overall wage or productivity level, as Carney is supposed to have claimed.

A 2014 Home Office report concluded that this effect on low-paid workers was found during recessions but not periods of economic growth, and was small and temporary in any case. The effect was not present in other government studies at all.

NIESR’s study into the impact of immigration on native British productivity found it to be small but positive. In general I suspect that there is a strong relationship between how good immigration is for natives and how flexible the receiving country’s labour market is.

The Mail is the Mail, but I can’t quite understand why the Times, in particular, decided to report Carney’s remarks in this way. To misinterpret him so badly seems to almost wilfully prefer a good story to an honest one.

Why we vote the way we vote

In my last post I tried to understand why people vote, suggesting that even if a sense of civic duty or a desire to express oneself can explain why we turn out to vote, these can’t really tell us much about why we vote the way we vote. In this post I’ll try to explain why I’m convinced that, for voters, ideas matter.

There are two basic views among political scientists about this: people vote to maximise their own wellbeing (“pocketbook” voters) or people vote to maximise the wellbeing of their society (“sociotropic” voters). The literature here is enormous so this post will try to sketch out the argument broadly – it is not intended to be anywhere near comprehensive.

There is a clear correlation between declines in GDP per capita and declines in support for the political party in power (‘economic voting’). But this could be because people who are worse off are changing their votes to improve their own welfare, or because people in general are trying to improve their society in general.

In ‘Sociotropic voting: The American case’, Donald Kinder and D. Roderick Kiewiet look at how voters behave when their personal circumstances differ from those of society in general – if you are unemployed, but total unemployment is low, are you more likely to want a change of government?

Looking at Congressional elections during the 1970s, they find strong evidence that people are more concerned with society and the economy as a whole than for their own circumstances.

‘A person’s private economic experience had very little impact on his choice of candidate in the congressional elections whereas his sociotropic judgements were of the utmost importance … American voters resemble the sociotropic ideal, responding to changes in general economic conditions.’

Kiewiet’s conclusion in a later book is that people blame factors other than the government for their own circumstances, but blame the government for the overall state of the economy. Is this a uniquely American phenomenon, though?

Leif Lewin’s review of the evidence in his excellent Self-interest and public interest in Western politics suggests that it is not – Western European voters, including British voters, also seem to be much more inclined to vote sociotropically than with regard to their own circumstances.

We know that voters are mostly very ignorant of the facts of politics, which may make it very hard for them to form accurate judgements about the best policies to achieve the end-goals they have in mind. But it also means that the media that they do pay attention to has an enormous influence over their perceptions, and that people’s political awareness may affect how ‘benevolent’ they really are.

In light of this, Gomez and Wilson (2001) adapt the pocketbook thesis to argue that more sophisticated, politically aware voters are more likely to be affected by pocketbook factors than others.

They are the ones who can think in terms of specific policies, make connections between particular policies and their own incomes, and do not blame incumbents for everything that goes wrong with the economy.

Other, less sophisticated voters simply assume that the President is responsible for what goes wrong with the economy. That might explain why electoral ‘giveaways’ (pensioner bonds, opposition to new home builds) seem to be concentrated on quite small groups of well-heeled voters – nobody else would notice.

The last word on voter behaviour must go to Philip Converse, whose 1956 survey data showed that most voters make their decisions based on extremely broad judgements of the ‘sign of the times’ (22%), or based on which group – posh people? workers? – a party or politician seems to speak for (45%), or even evaluations that had no shred of policy significance whatever, like which candidate was the funniest (17.5%).

Only around 15% of voters used ideology or ideology-like rules-of-thumb to decide who to vote for, and those were the most rigid in their decisions about how to vote.

To sum up, people seem to mostly vote for the candidates that they think will be best for society as a whole, though they may make very poorly considered judgements of that. If there is a ‘pocketbook’ effect, it is probably limited to the most well-informed voters.

All this suggests that the public choice view of democracy as just a way to divide the spoils of government between interest groups may well be wrong. Yes, voters are amazingly ignorant of basic facts, let alone economic theory, but we do have a chance of persuading them and changing the world for the better. To those of us who would like to believe in the power of ideas, that’s something to celebrate.