It used to be simpler back in the day. According to the nineteenth century aphorism, there were three kinds of lies: lies, damned lies, and statistics. But today, we also have economic forecasts to contend with.
The welter of revisions, upgrades and adjustments to the UK’s economic prospects in recent days, not least by the Bank of England today, illustrates just how central forecasting – the process of producing what we might call speculative statistics – is to modern political economy.
But does this make for good policy? After all, a lot of policymaking since Hunt’s November budget has been based on forecasts for falling inflation but also negative or at best flatlining GDP growth in 2023; instead, provisional data from Q1 now points towards stubbornly persistent inflation but a rather more resilient economy than generally anticipated (though with some worrying signs from the banking sector).
Part of the general problem with relying on forecasts is that the tail all too often wags the dog. Economic models can be built (and fudged) so that the desired output determines how inputs and variables are weighted, which assumptions are built in or left out and what levers are pulled. Anyone who has spent any length of time playing around in economic models or hunting for gremlins in the columns of OBR spreadsheets will know this to be true.
This isn’t always or even usually about a deliberate desire to mislead. More often than not, questionable inputs and assumptions are the accidental result of motivated reasoning, or missing information and time pressures, or the desire not to predict anything too outlandish – the same herding problem we’ve seen with political polls in recent years.
But sometimes, forecasts are manipulated intentionally, in order to produce outputs that align with an overtly political agenda. The most infamous example of this in recent years is probably the comically hysterical Treasury forecasts about the immediate impact of a vote for Brexit in 2016. In such cases, the intellectual integrity of forecasters is all too readily sacrificed in the pursuit of political priorities.
The high-profile uses and abuses of forecasts in political campaigns over recent years has thankfully engendered a degree of scepticism about the role they should play in policymaking. Wherever you sit on the political spectrum, you can probably think of several instances of slightly dodgy forecasts being used to confer spurious authority on an otherwise weak argument.
But in our justified scepticism, we should also be careful not to throw the baby out with the bathwater. Economic forecasting still has a role to play in policymaking, if approached in the right spirit. At the very least, if we approach the task with intellectual sincerity, the exercise of thinking about how to model a phenomenon or course of action forces us to consider the variables, factors and relationships at play. It highlights trade-offs, second order effects and probabilities. And it forces us to confront uncertainty, contingency and chance.
Economic forecasting can thus promote holistic thinking about problems in public policy. For example, if economic modelling had played as much as a role as epidemiological modelling in the UK’s Covid response, then the trade-offs inherent in the policies actually adopted – lockdowns, furlough, social distancing – would have been in sharper focus and perhaps better decisions would have been made during the pandemic, and maybe we would be in a better economic position now as a result.
Forecasts can also help us to look beyond the day-to-day imbroglio of Parliamentary politics and see the big picture. If we can identify structural trends that are likely to shape our economy and society over many decades, then in theory we have the chance to ready ourselves for future challenges. Many people would put climate change in this category. Likewise an ageing population, or mass migration.
The downside to taking the long view is that even a small change in the initial conditions or guiding assumptions can compound into massive differences by the end of say, a 50-year forecasting period. Demographic forecasts – crucial components within many broader economic forecasts – are notoriously hard to get right. We only have to look at China’s ageing population, which is partly the legacy of pro and then radically anti-natalist policies, to see the dangers of overcorrecting for future unknowns.
Yet on the whole, using forecasts to take a long view of things is a valuable intellectual exercise. While we should take overly precise figures with a pinch of salt, it can give us a sense of the direction of travel and prompt us into asking the right questions. It can also throw policy failings into relief.
For example, had GDP per capita growth reverted to the pre-2008 trendline rate after 2010, we would be almost 20% wealthier than we are right now. Thinking about the reasons for the vast gap between the reality and the logical counterfactual is arguably the single most important thing a think tanker could spend their time doing. So among other things, an economic model is a framework or heuristic that allows us to orient ourselves in a complex and confusing world.
Of course, we should reject positivist conceptions of social science and acknowledge that there will always be value judgements embedded in an economic model at some level. Similarly, in good Hayekian fashion, we must not lose sight of the fact that models are a simplification of reality – a poor proxy for the epistemic process and emergent properties of markets and other complex adaptive systems. Humility is called for, and a willingness to revisit past forecasts and reflect on what we got wrong. But more openness, scepticism and humility in economic policymaking would be no bad thing.
In any case, economic forecasts are inescapably part of the idiom of modern politics. As we have seen today with the Bank of England updates, they can shape the news agenda and set the political weather. Indeed, forecasts can have an almost totemic or talismanic power, especially insofar as political and economic agents treat them as authoritative. Scepticism about forecasts is one thing, but trying to escape the econometric paradigm altogether seems a futile exercise – the consequences of the Truss government’s side-lining of the OBR perfectly illustrates this point.
Just as lies, damned lies and statistics will always be part of politics and policymaking, so too will be economic forecasts. But approached with openness, sincerity and scepticism, that does not mean they cannot be useful devices in the policymaker’s toolbox.
Click here to subscribe to our daily briefing – the best pieces from CapX and across the web.
CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.
