In 2018, the Equality and Human Rights Commission commissioned and promoted a report which predicted that an extra 1.5 million children would be plunged into (relative) child poverty by 2021/22 if the government implemented Universal Credit. The proportion of children living in (relative) poverty would, it said, rise from 29 per cent to the unprecedented figure of 41.3 per cent.
Portes’ prediction was a totem for all economic forecasting, most of which is little better than guesswork and should not be taken seriously
If you think such prognostication is beyond the remit of the Equality and Human Rights Commission, I can only concur. The report was written by Howard Reed and Jonathan Portes. I know Reed’s work because he is often commissioned to produce research for nanny state groups. For example, he recently claimed that smoking costs Britain £173 billion a year! I know Portes a little, mostly through a series of increasingly ill-tempered Twitter interactions, and expressed to him my disbelief about his poverty prediction. This led to us making a substantial wager in 2018 which, after a little compromise, I agreed to pay him £1,000 if (relative) child poverty exceeded 37 per cent in 2021/22 and he agreed to pay me the same if it didn’t.
I didn’t read the report and probably never will. All I needed to do was look at the form. Between 1995 and 2016, relative child poverty (after housing costs) had fluctuated within the range of 27 per cent to 34 per cent. Since 2008, it had never gone above 30 per cent despite constant warnings about spiralling poverty under Tory ‘austerity’. For Portes’ prediction to be correct, child poverty would have to reach unprecedented heights within a relatively short period of time. It just didn’t ring true.
To my mind, an increase of this magnitude would require something more drastic than Theresa May’s reforms to the benefits system. Universal Credit was not implemented very well – it is a government programme after all – but it is not particularly stingy. There is too long a wait to collect your money once you start claiming and the monthly payments are not ideal for people who struggle with budgeting, but it tries to address the genuine problem of the poverty trap. My suspicion was – and is – that Portes’ prediction was based on a static model in which people do not respond to incentives and disincentives.
For as long as I can remember there have been reports from the likes of the Joseph Rowntree Foundation and Resolution Foundation predicting spiralling inequality and poverty unless the government reverses course on some policy or other. They are generally wrong, usually by a wide margin. In my more cynical moments, I wonder whether their authors know that they are wrong and only publish them to generate media coverage to put pressure on the government to maintain the status quo.
Portes accompanied his report with an article in the Guardian urging the government to ‘rethink’ its plan. So far, so normal. But then he took the bet, which led me to think that, unless he has money to burn, he must have really believed in his prediction.
If so, he was wrong. Very wrong. When the new figures were published this morning, they showed that relative child poverty (after housing costs) was still at 29 per cent, exactly the same level as it had been when we made the bet.
This humiliating defeat is particularly awkward for Professor Portes as he has spent the last few years telling me that I know nothing and he knows it all. As you can imagine, this is gratifying for me, not least because it wipes out gambling debts racked up during the World Cup in which I was badly let down by Belgium. My only regret is that I didn’t specify that the prize be adjusted for inflation. Thanks to the neo-Keynesian economic policies Professor Portes promotes, £1,000 in 2018 is only worth £828 today.
To the handful of lemon-suckers who say it is in bad taste to bet on child poverty, I say that we were not betting on child poverty. We were betting on the validity of economic forecasting. Portes’ prediction was a totem for all economic forecasting, most of which is little better than guesswork and should not be taken seriously. For example, in February, just four months after Liz Truss had to be restrained from ‘blowing’ a £30 billion ‘hole’ in the public finances, the Office for Budget Responsibility admitted that tax receipts were £30 billion higher than it had predicted.
As a rule of thumb, the greater the political implications of a prediction, the worse the prediction will be. We saw this with the Covid-19 models in 2021 and we see it with economic forecasting all the time. The pessimism – or, more rarely, optimism – of the authors spills over onto the spreadsheet. The reason the Covid-19 models were wrong is the same reason economic models are usually wrong. They are unable to predict how people behave when incentives change.
Besides, it was Portes, not me, who was banking on (relative) child poverty rising, and relative measures of poverty are not a good way of measuring deprivation anyway. Notice in the graph above that child poverty dipped between 2008 and 2011, and again in 2020/21. These were, to put it mildly, not good years for the economy and they were not good years for the poor either. Relative poverty fell because it is measured as the proportion of households living below 60 per cent of the median income. The standard of living did not improve for people on low incomes in those years. All that happened was the median income fell.
What we saw after the financial crisis is what we are seeing now: wages rising below the rate of inflation while benefits, for the most part, are lifted in line with inflation. If you used relative poverty as your guide, you might think things were getting better. In fact, the best most people can hope for is for things not to get much worse.
The same is broadly true of using income inequality as a measure of progress. Like relative poverty, it tends to go down in the bad times and up in the good times. The irony is that there has been plenty to be gloomy about in the last 15 years. Productivity, gross national income and wages have risen sluggishly and fitfully. The 2010s were not far off being a lost decade and the 2020s have been even worse so far. It has been a good time to be a doom-monger. The problem is that most of the doom-mongers have focused on the two metrics that have not got worse.
Update: Jonathan Portes has paid up.
Comments