Stumbling and Mumbling

Biased news

In 2010, 140,000 children aged under five died in Bangladesh. If the country had the same mortality rate (pdf) as the UK, only around 15,000 would have done so. This implies that around 125,000 Bangladeshi children die each year from poverty.

This fact, however, does not feature prominently in nightly news bulletins, even though it is equivalent to two Rana Plaza collapses every week.

There is, of course a simple reason for this. The news reports abnormal events, not normal ones; "dog bites man" is not news. Collapsing buildings are abnormal and so newsworthy whilst acute poverty is normal and so isn't news.

This bias is inherent in the nature of news. And yet it can be misleading. You cannot understand why so many Bangladeshis tolerate working in sweatshops until you realize that doing so gives their children not just a better chance in life, but a better chance of life. Thanks in part to the economic development brough by those sweatshops, child mortality in Bangladesh has fallen.

However, news reports which draw attention to the evils of sweatshops but not to those of rural poverty understate the benefits which such sweatshops have brought. Yes, they're hellholes which perhaps could and should be improved upon - but they're better than the alternative.

In this sense, news generates a bias amongst its western consumers; it encourages a hostility to globalization and industrialization even though these are - albeit imperfect - routes out of poverty.

There's a parallel here with attitudes towards crime reporting. It's a commonplace that whilst crime has fallen in recent years, the fear of it hasn't. A big reason for this, I suspect, is that violent crime - being abnormal - gets reported whilst folks living safely, being normal, does not. Ordinary reporting thus warps our perspective.

You cannot reasonably judge a probability distribution merely by looking at the far tail of it. But this is what the news invites us to do. 

There's another relevant bias here. Whilst under-reporting deaths from rural poverty the news is full of the doings of the rich and powerful. This too can have pernicious unintended effects. Laboratory experiments (pdf) have found that the mere act of communicating with others can induce them to behave more altruistically towards us. This implies that we are likely to be better-disposed towards the rich and powerful than we otherwise would be, and less well-disposed to the silent poverty-stricken billions. This too generates a bias towards tolerating poverty.

I say all this as a caveat to a common complaint. Everyone complains - with justification  - about bad, right-wing, dumbed-down linkbait journalism. But even when journalists are doing their jobs well, they are contributing to some unpleasant biases, by the very nature of what constitutes news. You cannot, rationally, base your political opinions in what your see in the news.

Benefits: efficiency vs legitimacy

The suggestion that wealthy pensioners should hand back benefits such as the winter fuel payment poses another question: why do pensioners get hypothecated benefits such as free TV licences, free bus travel or fuel payments? Wouldn't it be simpler to just raise pensions generally? And if some pensioners are too rich, just tax 'em.

The issue here is not simply one of administrative efficiency, but of basic welfare economics.

To see the point, take free bus travel. For those who uses buses, this equivalent to a cash handout equal to the cost of their bus tickets. But for those pensioners who don't use buses much a cash payment would be better. Net, the cash payment is better than free bus tickets. And free bus travel incurs another cost, as it can encourage pensioners who wouldn't otherwise use buses to do so, thus adding to congestion and fuel costs.

Econ 101, then, tells us that cash transfers are better than benefits in kind. So, why are pensioners paid ear-marked benefits instead?

One possibility is paternalistic. Winter fuel payments and free bus travel nudges pensioners to keep warm or to visit their family in a way that higher pensions alone would not. 

But I want to suggest another possibility. It arises from a tension in the very concept of government.

On the one hand, the function of government is to be impersonal - to lay down general laws that apply to everyone; this was emphasised by Hayek and is captured by old saws such as "justice is blind" and "the law is no respecter of persons." But on the other hand, this isn't sufficient to generate popular support - as its very impersonality creates a distance between government and citizens. Politicians who want to "connect" with voters will thus want to appear human and caring. And they do so more by making hypothecated payments, such as winter fuel payments, than by raising pensions generally.

There's an analogy here with the economics of Christmas. Joel Waldfogel has shown that giving gifts is less efficient (pdf) than giving cash, for much the same reason that benefits in kind are inferior to cash payments; at best, the gift gives us what we would have bought anyway, and at worst, we'd spend the £10 on something we'd like better than socks.

So, why do we give gifts rather than cash? It's because doing so seems more human, more thoughtful, more caring. In like fashion, winter fuel payments seem more caring than raising the basic pension.

If I'm right, we have a conflict between legitimacy and efficiency. Efficiency requires simple cash benefits, but these are cold-hearted compared to benefits in kind. I suspect this trade-off occurs in other fields. One reason why governments make interventions which libertarians (often rightly) think are inefficient - such as minimum wage laws or drug prohibitions - is that a government which concerned itself only with freedom and efficiency would be too cold and distant from voters to command much support.

And this in turn implies that a government of Hayekian classical liberalism, whatever its other merits, is just infeasible.

A complication: Winter fuel payments and free TV licences are fungible, in the sense that their recipients can use the extra income to spend on other things. This is why I call them hypothecated benefits rather than benefits in kind. I don't think this much affects my argument. 

Quantifying the evil of unemployment

Imagine an industry had, through its own recklessness, caused over 10,000 deaths in the UK. Or imagine that a government minister had, by his own stupidity, caused hundreds of deaths. There would surely be demands for severe punishment.

But in a sense, this is just what has happened, with pitifully little outcry. This is because, in utilitarian terms, the misery caused by mass unemployment is equivalent to very many deaths.

Let's start from the ONS's survey on the question: "overall, how satisfied are you with your life nowadays?" On a 0-10 scale, the unemployed report an average of 6.47 whilst the employed report an average of 7.53, If we assume the causality runs from employment status to well-being (so that being grumpy doesn't cause folk to lose their jobs), this means that unemployment costs the individual 1.06 "utils." On the same scale, married people score 7.72 whilst the widowed score 7.26 - so the death of a spouse costs 0.46 "utils."

Now, this doesn't mean unemployment is worse than the death of a partner. A person's death represents a loss to his other family and friends. Let's assume (maybe generously) that 20 others feel as bad as the widow. This gives us a total bereavement loss from death of 21 x 0.46 = 9.66 points.

In utilitarian terms, then, the death of a loved one is 9.1 times as bad as unemployment. But this isn't the end of the story. Death is permanent but unemployment isn't. Let's therefore weight the two events by duration.

Latest labour market flows data (pdf) show that during Q4 2012, 595,000 of 2.48m people moved from unemployment to work, which implies an average unemployment duration of just over a year. For simplicity, let's call it a year. How much longer does bereavement last?

Not long! Andrew Clark and Yannis Georgellis have estimated that there is "complete habituation" over time to widowhood; it takes an average of one year for men and three for women to return to their pre-bereavement level of well-being. Let's split the difference and assume two years of breavement. Let's assume it takes as long for family and friends to adjust. This gives us a total bereavement cost of 21 x 2 x -0.46 = 19.3 "utils".

To this, we must add the fact that death represents a loss of well-being to the deceased. Let's call this a drop 7.72 utils, the average for a married person. And let's assume, arbitrarily, that had our subject not died when he did, he'd have enjoyed another 10 years of this utility. This gives us a total utility cost of death of 96.5 utils.

This gives us a utility cost of death of 91 times that of unemployment.With unemployment now 944,000 higher that at the end of 2007, this implies that the welfare cost of recession is equivalent to the deaths of over 10,000 people.And if you think Osborne's misjudged fiscal policies have added more than around 24,000 to unemployment - and I suspect his critics would put the number far higher - then he is, in utilitarian terms, worse than Britain's worst serial killer.

Now, this calculation isn't intended to be remotely precise; it's a Fermi estimate. How should we respond to it? There are, at least, three possibilities:

1. It shows that there's something fishy about happiness research.

2. It shows that utilitarianism is wrong. Causing someone's unemployment is qualitatively different from causing someone's death - especially if you intend one but not the other - and you just cannot compare them on the same scale.

3. It shows that unemployment is indeed a social evil of huge magnitude, a fact which we don't realise because the jobless are socially and politically marginalized.

Personally, I'm not sure how to respond.

On wage-led growth

Yesterday, I quoted Michal Kalecki pointing out that wage cuts cannot create full employment. The fact that GDP has flatlined recently as real wages has fallen is consistent with this.This raises the question: could it be that policies aimed at raising wages at the expense of profits would boost the economy? A nice non-technical paper (pdf) by Marc Lavoie and Engelbert Stockhammer discusses the issues.

The case for such wage-led growth is straightforward. If the marginal propensity to spend out of wages is higher than that out of profits, a transfer of incomes from profits to wages will boost aggregate demand. This could happen. Common sense tells us that the badly off tend to spend their incomes, whilst the fact that rising profits before the recession were not accompanied by strongly rising capital spending suggests that the propensity to spend out of profits has been low.

This casual empiricism is supported by a paper by Ozlem Onaran and Giorgios Galanis, who have estimated (pdf) that, in many developed countries, rises in the share of profits in GDP have been associated with lower aggregate demand.

Better still, say Lavoie and Stockhammer, wage-led growth might be self-sustaining, in one of two ways:

 - Verdoorn's law tells us that higher output growth is usually accompanied by better productivity growth. This means higher wages might partly pay for themselves, and that they might not be very inflationary.

 - Firms might respond to higher labour costs by investing more in labour-saving technology, which would also tend to hold down inflation. If policy-makers respond to this by loosening macroeconomic policy, we could get a virtuous circle of non-inflationary demand growth, investment and rising productivity. This happened in much of the 50s and 60s.

So, what could go wrong? For me, there are three issues here:

1. A rise in wages for worse-off workers (for example through a living wage) would lead to lower payments of tax credits. In this sense, transfering money from capitalists to workers would be partly a fiscal tightening. That would tend to depress aggregate demand.

2. Is the marginal propensity to consume out of low wages high? If households are over-geared, they could use the extra income to reduce debts instead.

3. The relationship between capital spending and profits is unstable, as it depends upon animal spirits. There are two especial difficulties here:

- Capitalists would be happy to invest with low profit margins, if they are accompanied by high expected demand and hence high output-capital ratios and a high profit rate. But confidence that this will be the case requires them to have confidence both that macroeconomic policy can maintain high demand and/or that overseas demand will be high. This isn't certain.

- For the last 30 years, capitalists have been accustomed to policy being made in their own direct, short-term interests. Any disturbance to this might weaken animal spirits by more than the prospect of high demand would improve them.

Such instability is no mere theory. In the 50s and 60s, wage-led growth was feasible because capitalists invested in the confidence that high aggregate demand would be sustained. But in the 70s, this changed as they fretted about the squeeze on profit margins.

These doubts hint at what Servaas Storm calls a "sobering asymmetry". He says:

Lowering real wages will be unambiguously counterproductive, but the opposite policy of raising real wages will not have much of an impact on the economic recovery.

Perhaps the ability of capitalism to provide full employment requires freakish historical circumstances of a sort we saw only in the 50s and 60s. And perhaps it is only wishful thinking by both rightists and social democrats that stops them seeing this.

Unemployment & the free market

Bryan Caplan deserves praise for calling on free market economists to pay more attention to the "grave evil" of unemployment. I fear, though, that he overstates what free market policies can contribute to solving the problem.

My chart shows the problem. It shows the UK unemployment rate between 1855 (when data begins) and 1914. You can see that the jobless rate was often high - it averaged 4% - and volatile.

And this was during a period of as free markets as one could practically get.

This undermines at least three "free market" explanations for unemployment:

- "Welfare benefits mean the unemployed have little incentive to get work." In the 19th C, though, the only state support the unemployed got was in the Workhouse - and even as late as in my lifetime, this was spoken of with terror.

- "Big government and taxes deter job creation." But public spending in this time averaged only around 10% of GDP, and labour market regulation except for a few Factory Acts was nugatory by modern standards.

- "Wages are too rigid". But wages fell in nominal terms in 13 of the 59 years here, and in real terms in 12 of these years. Average nominal wages fell by 9% between 1874 and 1879, which is consistent with some sectors seeing very large falls.

There is, though, an alternative theory that fits these data. It's that a free market will see large swings in aggregate demand and employment, and that unemployment cannot be prevented by wage reductions alone. This was pointed out most famously - well famous in my house anyway - by Michal Kalecki in 1935:

One of the main features of the capitalist system is the fact that what is to the advantage of a single entrepreneur does not necessarily benefit all entrepreneurs as a class. If one entrepreneur reduces wages he is able ceteris paribus to expand production; but once all entrepreneurs do the same thing - the result will be entirely different.
Let us assume that wages have been in fact generally reduced...and in consequence unemployment vanishes. Has depression thus been overcome? By no means, as the goods produced have still to be sold...A precondition for an equilibrium at this new higher level is that the part of production which is not consumed by workers or by civil servants should be acquired by capitalists for their increased profits;in other words, the capitalists must spend immediately all their additional profits on consumption or investment. It is however most unlikely that this should happen...It is true that increased profitability stimulates investment but this stimulus will not work right away since the entrepreneurs will temporise until they are convinced that higher profitability is going to last...A reduction of wages does not constitute a way out of depression, because the gains are not used immediately by the capitalists for purchase of investment goods. (Selected Essays on the Dynamics of the Capitalist Economy, p26-28).

There's a good reason why almost all major economies abandoned free market economics. It's that such economies didn't and couldn't avoid mass unemployment.

I'll concede - much more than most lefties - that there's a big place for free market economics. But the labour market ain't it. 

Note: data comes from the Bank of England.

Invisible enemies

Jason Cowley laments the lack today of literary figures who can write seriously about politics.

I don't think this absence is wholly to be regretted; fine writing can easily become mere egomaniac poncery. But there is, I suspect, one reason for it which is largely overlooked. It's that the political enemies today are invisible ones, which don't lend themselves easily to cultural depiction.

To see my point, contrast the (rationalized) Thatcherite analysis of the UK's problems in the 70s and 80s to now. Back then, the "enemy within" was in plain sight. It was unions. Smash these, it was thought, and you'll increase profit margins and management's "right to manage", and this will increase business confidence and hence investment. This worked, sort of, at least for a while.

Today, though, we can't blame a bogeyman for our economic problems. The lack of investment opportunities, slowdown (pdf) in technical progress, collapse in demand for unskilled labour, inability of capitalism to create full and meaningful work, and ideologies that sustain managerialism and inequality are all impersonal problems. You can't solve them by defeating a visible enemy.

In this context, it's no surprise that literature and culture generally have become disengaged from politics. It's possible to make good art about corruption, racism, warmongering or recession. But try doing so about the investment dearth.

Herein, though, lies a problem. Very many people active in politics don't get this. They are still trying to fight a bogeyman - be it immigrants, "scroungers", greedy bankers, or tax-dodging companies. In this sense, it's not just literary figures who are unable to address our political-economic problems. Political figures can't do so either.

On disgust

Some of you might find it distasteful to make a link between the Boston bombings and Luis Suarez. But this is precisely the point, because there is indeed a link, and it highlights a difference between two different moral perspectives.

In Suarez's case, many pundits are calling for him to be banned for a long time for biting Branislav Ivanovic.This means he would get a longer ban for doing something which didn't injure his opponent than the three-game bans which are usually given for the sort of reckless tackles that can lead to serious injury. Many might think this inconsistent.

And Michael Cohen points out that Americans' responses to the Boston bombings are also inconsistent. He notes that 11 people were killed by guns on the day of the bombings alone, and that, over a year, thousands more Americans are killed by gun crime than by terrorism. And yet Americans are more willing to sacrifice freedom and money to fight terrorism than gun crime.

There's a common theme here. It's that many people are not utilitarians. Their moral sentiments are based not just upon objective costs and benefits but upon a feeling of disgust. Sure, Suarez imposed less injury upon his opponent than many reckless tackles do. But biting someone is disgusting whilst fouling them is not, and so many feel it deserves greater punishment. Similarly, terrorism evokes disgust to a greater extent than does gun crime, and so justifies a stronger policy response even though the objective costs of gun crime are greater.

This disgust has economic implications because, as Alvin Roth has described (pdf), it can prevent otherwise mutually beneficial trades such as in markets for donor organs.

In this sense, there is a direct conflict between the utilitarianism that is the default position of economists and public opinion. Many economists say "markets facilitate mutually beneficial trade, so we should have free(ish) ones in organs, drugs, prostitution etc." To which non-economists go "yuck, that's disgusting."

Herein lies a reason why, historically, utilitarianism and liberalism have tended to go together - most notably in the person of John Stuart Mill. In considering only objective costs and benefits, utilitarianism discounts disgust. It thus removes a justification for criminalizing "self-regarding" acts such as prostitution, drug-taking or (for many years) homosexuality.

And in this last word lies a problem. Over time, the things that disgust us change. Today, we find the slave trade disgusting but homosexuality not. But there was a time when the opposite was the case. Granted, there seems to be a neurological basis for the feeling of disgust - but the objects which trigger the feeling change over time.

Which raises a problem. Does this show that feelings of disgust are irrational prejudices? If so, should they have weight in policy-making? Or are they instead forms of taboo or social norms which are rationally defensible? If they are, how can utilitarians and liberals better engage with them?

I have no answers here. I'm merely pointing out that we have here two competing moral perspectives, and very often the clash between the two produces merely a sterile slanging match.

Another thing: I'm ignoring here the fact that Suarez has lengthy form and so reactions to his bite are shaped by a form of framing effect.

What if demand for graduates falls?

What is university for? I ask this old question because the utilitarian answer which was especially popular in the New Labour years - that the economy needs more graduates - might be becoming less plausible. A new paper by Paul Beaudry and colleagues says (pdf) there has been a "great reversal" in the demand for high cognitive skills in the US since around 2000, and the BLS forecasts that the fastest-growing occupations between now and 2020 will be mostly traditionally non-graduate ones, such as care assistants, fast food workers and truck drivers; Allister Heath thinks a similar thing might be true for the UK.

There are signs of this happening already. ONS data show that employment in professional occupations, having grown 3.1% a year between 2001 and 2007, has grown just 1.2% per year since, whilst overall employment growth hasn't changed much.

And we can tell a futurological story in which high-end work declines: IT could replace accountants and lawyers, not just routine clerical workers; MOOCs could render some college lecturers redundant; a smaller financial and public sector will limit demand for graduates; and the increased supply of graduates from India and China might bid down the wages of their western counterparts.

Now, none of this is at all certain. In 1975 Gary Becker, the founder of human capital theory wrote: "Perhaps the current weak demand for highly skilled manpower is the beginning of a resumption of the earlier [1900-40] decline." (Human Capital, 3rd ed, p9) That was just before a two-decade long increase in demand for graduates.Which tells us that forecasting demand for skills is a mug's game.

Nevertheless,we should ask: what function would universities serve in an economy where demand for higher cognitive skills is declining? There are many possibilities:

- A signaling device. A degree tells prospective employers that its holder is intelligent, hard-working and moderately conventional - all attractive qualities.

- Network effects. University teaches you to associate with the sort of people who might have good jobs in future, and might give you the contacts to get such jobs later.

- A lottery ticket.A degree doesn't guarantee getting a good job. But without one, you have no chance.

- Flexibility. A graduate can stack shelves, and might be more attractive as a shelf-stacker than a non-graduate. Beaudry and colleagues decribe how the falling demand for graduates has caused graduates to displace non-graduates in less skilled jobs.

- Maturation & hidden unemployment. 21-year-olds are more employable than 18-year-olds, simply because they are three years less foolish. In this sense, university lets people pass time without showing up in the unemployment data.

- Consumption benefits. University is a less unpleasant way of spending three years than work. And it can provide a stock of consumption capital which improves the quality of our future leisure. By far the most important thing I learnt at Oxford was a love of Hank Williams and Leonard Cohen.

But are these benefits really worth £27,000. I fear that, in an economy with declining demand for graduates, many will think not. If so, rather than supply much-needed educated workers, universities might merely create a (larger!) mass of disaffected 20-somethings. In this sense, the technical-economic matter of relative demand for skills could have interesting social consequences.

Another thing. It's possible that a society of educated people is likely to be more cultured and scientific-minded than one of non-graduates, and this should have positive externalities in the form of better political discourse and higher culture. There is, however, little evidence of this in practice.

Endogenous morality

Recent events pose a challenge to conventional neoclassical economics. I refer not to the financial crisis, but to a fat Geordie's attempt to twat a police horse.

This seems to undermine the standard economics of crime, as developed (pdf) by Gary Becker. This says that people will commit crimes if the expected benefits outweigh the costs. Such a theory, however, doesn't fit two big facts. One is that many crimes, such as twatting p.hs, are just moments of madness. The other is that many people eschew crimes even though they might well get away with them. These two facts can co-exist in the same person; if we take our fat Geordie at his word, he has "never been in trouble before."

However, a new paper by Matteo Cervellati and Paolo Vanin of the University of Bologna helps to explain such behaviour.

They start by recognizing weakness of will, that we are tempted to do things we'll regret. This introduces a role for morality. It works as a self-commitment device; "thou shalt not steal" stops us breaking into people's houses if we see the window open.

Such moral codes can explain our fat Geordie's behaviour. Whilst they can stop us committing otherwise profitable crimes, they don't always prevent us being stupid in the heat of the moment.

So far, so trivial, you might think. No so. Morality has a cost; it can either prevent us committing profitable crimes or it can impose a burden of guilt upon us if we do so. It follows that parents who care for their children's material welfare will not always want to inculcate into them a strong moral code. If you think your children will be so poor they will starve to death unless they steal food, you'll be less keen than others to tell them that theft is wrong. Or if you think your children will be so rich they can do what they like, there'll be no point burdening them with guilt.

Cost-benefit considerations, therefore, tell us that moral codes will be strongest among what used to be called the middling sorts. Parents who feel their children will be sufficiently poor that they'll be tempted to steal, but also sufficiently well-off that they have something to lose, will be most keen to inculcate morality.

This fits historical stereotypes; the feckless underclass, the dissolute aristocrat, the respectable working class man and the priggish bourgeois are all cliches. But they are cliches for a reason. Cervellati and Vanin use people's attitudes to tax-dodging, fare evasion and benefit fraud in the World Values Survey to show that, indeed, middling classes do have a stronger moral code.

We might add that this also explains why bourgeois morality has so often been accused of hypocrisy. It's because such morality was partly cultivated by self-interest. (I nearly wrote "arose out of" but that pitches it too strong.)

Cynics might think all this is yet another exercise in economic imperialism - an attempt to explain everything with the tools of costs and benefits. But it is hinting at something which might be both true and important - that social norms and moral codes are not (just) exogenous constraints, but are instead economically endogenous. In this sense, neoclassical economics and Marxism agree.

Reinhart & Rogoff: true Keynesians

Any Keynesians who are rejoicing in the debunking of Ken Rogoff's and Carmen Reinhart's claim that high government debt reduces GDP growth should stop. I suspect Reinhart and Rogoff are, in fact, guilty of a Keynesian error.

To recap, they claimed (pdf) that ratios of government debt to GDP over more than 90% are associated with significantly lower average GDP growth. However, when Thomas Herndon, Michael Ash and Robert Pollin tried to replicate this result, they found that it was due largely to excluding Australian, Canadian and New Zealand experience in the late 40s, and an Excel error - which Reinhart and Rogoff admit was a "significant lapse"  - which led to good Belgian growth being excluded.Fixing these errors shows that high government debt has been compatible with decent growth.

So, why do I say Reinhart and Rogoff are Keynesians? Simple. The issue here is not really one of policy; insofar as Reinhart and Rogoff's result has been used by policy-makers, it is in the way that drunks use lamp-posts - as support rather than illumination.

Instead, the issue is about the culture of economics. And here, Keynes (among others) had an unfortunate if perhaps inadvertent effect. 

What leaps out of his General Theory is that it is entirely unencumbered by empirical evidence. Keynes thus helped to promote an ideal of the economist as a brilliant man capable of solving problems from his armchair by dint of superb intellect*. What got undervalued in this was the importance of the mundane grunt work of careful fact-gathering.

Reinhart and Rogoff's errors, I suspect, reflect a culture which prizes brilliance - and no-one doubts that Rogoff is brilliant - over dull pedantry.

When I started work, I realized that the job of the practical economist was not so much about theorizing but more concerned with gathering and understanding data - something which academe had wholly unprepared me for. For me, the difference between the professional economist and the amateur is that the former knows the numbers not in the sense of understanding high econometric theory, but in the sense of knowing where they are, what they mean and what they don't. This is no small skill. An ability to navigate the ONS website without recourse to language you wouldn't use in front of your mother requires a mastery of the arcane which is not given to many mortals.

I suspect things have changed a little since I was a student; a lot of good recent work in economics has involved using big data or generating facts by experiment - for example in the work of Mark Grinblatt, Mattias Sutter or Esther Duflo to name but three of many. But I'm not sure it's changing enough. Diane Coyle has reported that many employers want economics students to have "a better practical grasp of quantitative methods including collecting and understanding data (as opposed to more sophisticated econometric techniques)." Perhaps, then, the tendency to elevate the brilliant theorist over the empiricist lingers.

Traditionally, many economists have tried to model themselves on Isaac Newton - albeit emulating his autistic misanthropy better than his genius. But perhaps Darwin would be a better model. His greatness lay not so much in the theory of evolution - something quite trivial he took from economics - but in the years of careful fact-gathering that preceded it. Perhaps if this were so, the sort of mistake made by Reinhart and Rogoff wouldn't have happened.

* There's a contrast here between Keynes and Kalecki; the latter did develop hypotheses which he tested against data, which is is why I hold him in higher esteem than Keynes.