Ten mistakes that permeated the Channel 4 ‘How Rich Are You?’ programme on inequality (part 1)
1) Getting confused between wealth and income
The show constantly conflated, and moved seamlessly between talking about, income and wealth. In the very first minute of the show, the presenter Richard Bacon spoke about ‘wealth extremes’ and explained to viewers that the show would allow them to find out where they stood. But that whole first part of the show was about asking people where they stood in the distribution of income. Indeed, the very first pie chart shown on screen claimed ‘only 2% of people correctly estimated their wealth’, when it was clear again that what was being discussed was where people stood on the income spectrum. This sort change of the actual topic under discussion occurred throughout the show, making it difficult to follow exactly what the problem was that we were all there to talk about.
2) That wealth is a fixed pie
Whilst I have no doubt the people who produced the show understand that economic growth can mean everyone can become better off over time, much of the early discussion on inequality was presented as if the stock of wealth or even overall income was fixed. This was done through lots of loaded language, like the rich ‘siphon off money’ etc., and at times assertions that bordered on saying the rich were only rich because of exploitation that made the poor poorer. This is, to put it lightly, historically ignorant. To pluck an example from Deirdre McCloskey’s recent book, the average Swede in 1800 made the equivalent of $3 a day – now it’s as high as $120 a day. That hasn’t come about because of redistribution or bargaining for a bigger piece of a fixed pie, but because of the dynamic gains coming from innovative capitalism. The irony here is that McCloskey’s work concludes that it was a shift in the ideas away from the kind of anti-wealth narrative which dominated the show that facilitated the great enrichment we have seen over the last two centuries.
3) That poverty and inequality are interchangeable
Another theme that ran through much of the show was the interchangeable discussion of poverty and inequality. There was lots of discussion of people struggling to get by on low incomes – contrasted with those doing well at the top. But, again, there is little evidence that the rich being rich has caused the poverty of the poor. In fact, in recent years poor productivity performance coupled with a very high cost of living have been the main reasons why living standards have worsened. But attempting to do something about these problems would lead you down a very different policy path to one in which you were concerned primarily with the gap between rich and poor.
There may well be policy changes that could reduce poverty AND inequality at the same time. Liberalising our planning laws, for example, could reduce house prices. But we would want to do these because they are good for the economy and pro-poor – as opposed to because they might reduce the income gap. Unfortunately, rather than discuss how productivity growth could be enhanced or the cost of living reduced, much of the show simply discussed undesirable outcomes that we might expect to be associated with low incomes – and laid the blame firmly at the door of inequality, calling for yet more redistribution as a solution rather than broader growth.
4) That inequality is in and of itself a serious problem
The question that John Cochrane has asked on this blog before is: ‘Why is inequality a problem in and of itself, rather than representing a symptom of problems that should be fixed for their own sake?’ I’ve re-watched the show a few times, but am yet to find a convincing answer. Sure, some activities may lead to both more inequality and worse living conditions for all. It may be that educational inequalities – due to lousy state schooling in some areas – lead to diverging outcomes. But we can deal with these problems at source. Once you strip out these reasons for inequality, often the fault of existing policies, the real questions that egalitarians have to answer are: why does Bill Gates getting rich from a new Microsoft product (which may increase inequality) harm the poor? What’s the transmission mechanism? In other words, why does some people earning much more than others matter?
5) That Spirit Level style analysis is slam dunk evidence there is a transmission mechanism
One section of the show echoed the infamous Spirit Level (SL) with evidence that inequality was responsible for a range of social ills – essentially with correlations showing that if the gap is bigger we are more likely to: die earlier, marry poorer people and have bad educational attainment. It should be pretty obvious that most of these things are determined by absolute deprivation, assortative mating at universities and poor state schooling in many areas – not the ‘gap’ between rich and poor per se. They may correlate with the gap, but this is not the same as explanatory power, and so is not a useful guide to policy. It’s for this reason that the Spirit Level, which was based on lots of these correlations, has been labelled as ‘pseudoscientific nonsense’, and is not regarded as important work by serious academics in this area. To take just one critique, my colleague Christopher Snowdon has shown (using the same data sources as Wilson and Pickett) that ‘the results have been influenced by the choice of countries and indicators. Including just a handful of additional countries is enough to make many of the SL graphs, which show data points scattering around a straight line, dissolve into shapeless point clouds.’ In short, their analysis – the type of analysis at the heart of the show’s approach – amounts to a load of scatter plots which show links between inequality and a range of negative indicators but fails to account for endogenous or control variables, or reverse causation, with the analysis highly influenced by the selection of countries and time periods. To say this level of analysis is not high quality or robust would be an understatement.
6) Dodgy and misleading statistics on the distribution of income and wealth
One of the key parts of the show entailed asking people how much they earned, and where they thought they were on the distribution of income. This was clearly designed to show that as a country, we are poorer than we think we are (so individuals found themselves much higher than they were expecting!). For example, if I were to answer the quiz, I would find myself just inside the top 15 per cent of earners – though I feel nowhere near the top of the income pile. This is designed to make me shocked about the low incomes received by many people in the country – and presumably, how small the group is who earn significantly more than me.
But this crude comparison of where we are on the income scale misses a few obvious realities. First of all, most of us would tend to unwittingly judge ourselves in comparison to our contemporaries – i.e. our age cohort – and to people with similar work patterns to us. It is quite difficult conceptually to judge where you on the income spectrum in relation to someone working part-time or, say, just a few hours a week on a Saturday. There is also a clear lifecycle effect such that as you get older, you are likely to move up the income scale too.
Perhaps more important than this comparison effect though is the fact that the data does not take into consideration living costs, pensions provision, household composition, or the existence of the welfare state. If you live in London, then high rent or housing costs might mean that although your earnings put you in a high income bracket, your real disposable income after essentials is much lower. If you have lots of dependents, then your position on a household basis may be much lower. If you are contributing lots into a pension to smooth your living standards including your retirement, then this will reduce your effective disposable income and make you feel poorer in income terms than you actually are (particularly now we have a flat-rate state pension). And of course, what ultimately matters for most people is their living standards after taxes and benefits – ignoring this means not informing people that we already have significant redistribution by the state. Recent ONS analysis on households shows the average income of the richest fifth of households is fifteen times higher than the poorest fifth before taxes and benefits – but this falls to just four times more afterwards. It would seem pretty important for a programme lamenting huge inequality and with many guests calling for much higher redistribution to mention this.
Perhaps more misleading still, however, were some of the statistics used on wealth – including the use of Oxfam’s ‘just 5 families have as much wealth as the poorest 20 per cent of the population’. This sort of measure has been highly criticised because it uses net wealth figures – which imply people who may have very high incomes and significant assets but also have significant debts (such as large mortgages) would be defined as ‘poor’. In other words, Oxfam’s methodology means that a child with £1 would be wealthier than thousands of people with household debts. The effect of this methodology (adding negative net wealth values up) means that in order to get to a high total wealth number for those at the bottom end you have to move a long way up the distribution – giving us the meaningless statistic quoted above. Judging the wealth of the poor in this way makes no sense.
Mistakes 7-10 will be discussed in Part 2.