Egged on by politicians with an axe to grind, there’s a growing angry mob at the gates of many of the UK’s economic forecasting institutions.
Jacob Rees-Mogg recently hit the headlines for accusing Her Majesty’s Treasury of “fiddling the figures” in a leaked analysis of the possible effects of Brexit.
But while it’s tempting to see this as part of an ideological battle between the hidebound forces of civil service orthodoxy and the freewheeling iconoclasts of the libertarian fringe, what is really going on is a fundamental misunderstanding of the point of economic forecasting.
It is true that making econometric projections is a lot like trying to hit the bullseye on a randomly moving dartboard. You know the rough parameters of where you need to throw the dart, but you can only make a best guess of where, and how fast, the board might move.
Historic inflation, unemployment, and spending data may well help to deduce how GDP may perform in the years ahead, but ultimately it’s just a guess.
Of course, the economy is far more complicated than just prices, workers, and households.
Making forecasts means estimating the probabilities and weighing the likely impact of political, social, technological, and even security risks, alongside making rational assumptions about the irrational behaviour of economic agents.
As such, even with the promises of artifical intelligence and big data, forecasts will always be indicative – and should not be treated as gospel.
If the economy confounds expectations, that’s surely a good thing, and not necessarily just a chance to castigate forecasters with an “I told you so”.
This is because the Office for Budget Responsibility, HM Treasury, and independent forecasters can only really base projections on current government policy, and often unreliable clues on how it may evolve.
Conversely, gloomy projections should spur policymakers into growth-enhancing policies, which would not be possible for forecasters to pre-empt.
Forecasts are nonetheless an important benchmark for businesses to plan hiring cycles, product launches, and investments, and for policymakers to make appropriate fiscal, monetary, and supply-side interventions.
But even when we’ve fully accepted that forecasting is a non-exact science, how do we move forward from the current pall of scepticism surrounding the industry?
First, we need to recognise that having a wide breadth of forecasters gives us a much more rounded picture of the various perspectives on risk in the economy.
Individuals – politicians or otherwise – have a tendency to lean on judgements which reinforce their own beliefs or support a narrative they’re attempting to sell.
As such, it’s important to take notice of the vast swathe of independent projections, rather than focusing on a few.
However, it’s not just their bottom-line numbers we should be interested in. It’s their underlying assumptions – and the weight given to them – that are just as important.
While accuracy rankings exist for economic forecasters, which compare their projections against actual out-turns, it’s equally important to score them on the clarity of their models and assumptions.
Understanding the mechanisms and logic behind an institution’s conclusions is crucial, especially given this is the point at which forecasters are most likely to diverge.
Public and private forecasters must be more transparent about their methodologies, suppositions, and confidence intervals. The media must also help by balancing the need to make headlines by also scrutinising forecasters underlying assumptions.
It’s important to acknowledge the crucial role of policymakers in accurate forecasting.
Given the desire to influence economic behaviour in specific time horizons, the government cannot always be upfront about policies.
But clear and concrete policy announcements – rather than statements of intent – are crucial for economists and businesses to make judgements and decisions on the future.
So if MPs scold forecasters for getting it wrong, it may in fact be partially reflective of their own lack of clarity and conviction on policy.
Sure, economic forecasting may make astrology look respectable, as economist John Kenneth Galbraith once said. But without it, we would be walking forwards without a flashlight.
Rather than simply attacking forecasters for being wrong, we need to get a better grip of their limitations.