There’s an old saying in philosophy, as follows. There are only two answers possible to any given philosophical question: maybe, or, no one can know. Given that philosophy is about the nature of things, you might suppose this would be a description of our condition more generally. When things are complicated, we can only make guesses (even if really good guesses), or accept the limits to our understanding.
This is an insight that seems to have been lost in modern economics. The idea of limits to what we can say we know sometimes appears to have been abandoned. Despite the scientific methods of modern economics, at times there appears to be an absence of scientific doubt.
Perhaps it is because economics and politics are so closely intertwined. The political benefits accrued from attaching yourself to the economic consensus of the day are legion. Having credibility amongst academia, and therefore business, and therefore the public who listens to both, has electoral consequences. The economy is so often cited by the public as their primary issue of concern. So getting economic policy ‘right’ is the bedrock of creating an electable policy platform.
So far, so straightforward, then. You can’t run the country unless you can run the economy. And to run the economy you need to show how you would push or pull the levers available to generate better results. Unfortunately, this is where the politicians’ rush to economic certainty is at risk of ignoring constraints inherent in the dismal science. There are two important functions economist perform to assist policy makers. Firstly, there is the collection and interrogation of historic data to discover patterns and stories about how economies function. Secondly, there is the use of these insights, coupled with theoretical assumptions, to make economic models of future performance.
And this is where the row about the Thomas Piketty’s book on wealth has created a storm across the pages of broadsheets and blogs perhaps underestimates the limits to certain kinds of knowledge. The FT’s Chris Giles wrote this account of the mistakes he says Piketty makes. Branko Milanovic amongst others responded.
Part of what Giles says is that there are transcription mistakes; particular data sets incorrect because of slip ups. That’s surely a fair enough criticism. But some of what he points to is measurement issues. And I would have little criticism of the points he makes. What I would point out is the unwillingness of some to recognise that there might be no complete, definitive, answer to the distribution of wealth in the past. We cannot travel back in time and ask public officials to ask different questions. As Giles puts it in his article, “While this post is clear about what is wrong with Piketty’s charts, it is much less certain about the truth.”
Politicians who, like minister Matt Hancock (“Ed Miliband based his whole anti-business approach on a French economist comprehensively demolished by Chris Giles”), seize on Giles’s article as some kind of knockdown demolition of arguments against inequality underestimate the complexity of the task in hand.
It is good that writers, and serious economists, debate, question, and doubt. It is not right for politicos, in search of a quick win, leap on the conclusions of one economist, as if the question is answered, now and for all time. We do not act with intelligence and integrity if we do so.
In much more limited way, over the past year, colleagues and I have fought a battle with ministers and the ONS to correct or improve the questions in the Labour Force Survey that measure the use of zero-hours contracts. We knew because of data inconsistencies that there was likely more use of such contracts than appeared from ONS estimates. So we needed alterations to the data collection that would get closer to the real picture.
But in changing the survey’s methodology, the later information cannot genuinely be compared to data collected in the past. There may have been a sharp increase. The number may have been more than we thought all along. I may have my own guess. Others could estimate. No one can actually know, now. Likewise, there may be matters concerning inequality about which we should conclude it is unlikely we’ll ever know the answer.
On the forecasting side there are issues too. An economic forecast is a set of assumptions, married with data sets capturing the current economic position, to give a description of likely future outcomes. “Likely” is an important side-constraint here. In the end, despite the mathematical validity of the structure of the model, it’s a human being who has established the model’s assumptions, and weighted the evidence, that drive its predictions.
There is lots of writing about this, but very recently Steve Van Riel has described some of the prediction biases the human mind can bend towards. Furthermore, models suffer in exactly the same manner as historic analyses given the issues with data collection described above. Our ability to predict the future depends on our understanding of what happened in the past. And as we’ve seen, what happened before isn’t always exactly clear.
George Osborne perhaps nodded to some of these issues in the creation of an independent Office for Budget Responsibility. And that’s why Labour has suggested that the OBR independently audit the costings of spending and tax commitments made by all major parties. However, whatever the value of independence, in itself it doesn’t make the model any better. The forecast is still beset by the same constraints as any other, whether independent of ministerial department or not.
So how should politics respond? In short: humility. Why is it so unacceptable to say that we don’t always know the answer? Matthew Taylor has written about open policymaking that acknowledges that some courses of action will be uncertain, and may need to alter as more information and evidence becomes available. Of course, this might appear to be problematic. Politics is about making bold claims. But I maintain it is not good politics to go around making unscientific claims. In so far as the rigour of science can be brought to bear on economics, we need to express much more scientific doubt and accept the possibility of new information, new data, or a new picture emerging.
Our boldest claims should be reserved not for perceived proof of our point against our political enemy, but for the values we hold. Let the public know what we stand for, and for many of us, that will include a moral argument against inequality. Especially the kind of extreme inequality that leaves some in slums whilst others reside in great comfort. Or inequality that means the talents of some are unlikely ever to be realised, frustrating the ambition of those who happen to be born poor, and depriving the wealthy of the contribution that could be made those not so fortunate.
Whilst there may be compelling economic reasons that such inequality is bad for everyone, rich or poor, to be lead by the evidence means remaining open to new data that demonstrates the opposite. To hold moral values requires no such commitment. It is a deep seated value in so many of us that people, underneath it all, are fundamentally the same, of the same worth and dignity. Which is why extreme inequality is wrong, and cannot ever be right. This cannot be altered by a new data set, or next year’s survey. It is true as long as we are all human.