Review: A New Understanding of Poverty
Liberals frequently find themselves in the middle of debates. That on poverty is little different. What is poverty and how it is measured is an academic discussion, but one that has far reaching consequences for welfare policy and those impacted. The Institute for Economic Affairs, the leading free market think tank, has examined this issue through Poverty Fellow Kristian Niemietz’s new monograph “A New Understanding of Poverty” .
On the wings of this discussion are the two best understood concepts, absolute and relative poverty. Preference for those measures tends to ascribe to the right and left wings of politics respectively.
The study shows that different poverty measures are poorly correlated. Countries that have seen large rises in relative poverty recently have seen sharp falls in the absolute. Even different relative measures are more weakly linked than we might assume. It is then incorrect to assume different measures are just alternate views of the same thing. What measures are used to assess poverty then matters.
Absolute poverty, the measurement of access to some unchanging basic necessities such as water, food, and shelter was popular in the Victorian era (for example the Rowntree Budget Standard Approach) and accurately describes subsistence, but doesn’t adapt to progress. Access to a phone 50 years ago was a luxury, today it is essential. The poor in Britain today are better off than the middle classes in the 1950s. Applying Victorian or third world standards to Britain today tells us mainly that the Northumbia students who compared their tuition fees plight to famine victims last November are getting a poor return on their investment in education.
Relative poverty, for example setting a poverty line at 60% of median income, moves with economic changes, but it does so perversely. RP for example falls in recessions as relatively more expensive employees are laid off. It can increase as everyone gets better off if that growth is uneven. In uncertain times someone might find themselves labelled poor or not poor several times over without any actual change in their standard of living. It depends entirely on what geography is selected.
RP has little relation to what most people think is poverty and breeds resentment of the welfare system. By focusing on income it labels fabulously asset-rich pensioners, future well off students, and temporarily unemployed professionals as poor. Targeting RP leads to policies focusing on the line and shuffling money about not tackling the root causes of deprivation. By making a fetish of redistribution RP targets can encourage high taxes and welfare spending that destroy growth and makes joblessness a lifestyle choice (under socialism we all get poorer together, but none more so than the genuinely deprived).
Targeting inequality as proposed by the debunked pseudo-science of the Spirit Level doesn’t work. Current UK policies aspiring to abolish relative child poverty and other targets, will either never be achieved, or rely on economic collapse.
Subjective poverty measures are usually survey based. Either percentages self-reporting ‘I feel poor’ or based on an income/wealth line where that becomes more true than not. Such emotive measures are interesting studies in self-perception, but don’t tell us much about how deluded those perceptions are or why they matter. As a policy tool they have much of the utility and of an ‘Am I fat’ survey in Heat Magazine. A policy prescription to tackle subjective poverty might include a National Hugs Service*.
Material Deprivation measures are based on reported expenditure on goods and services, weighted by how important they are to the individual, and how widely they are owned (consensual approach). The result is a range of context specfic deprivation scores rather than a poverty line. Expenditure measures move with economic progress, correlate with popular perceptions of poverty, and can account for spending from assets. The survey element is binary – do you have this, don’t you; do you want it, don’t you – which should mitigate against perception bias.
Another advantage of MD measures is that they improve when essentials get cheaper, usually as a result of more competition or innovation. The deteriorate when the converse is true. Government schemes that increase the price of rent, petrol, electricity, food and clothes for example, and the subsequent household spending adjustments are better captured by an MD scale than RP. They can account for regional differences in price.
The main problem with MD is preference bias. Some who forgo ‘basic’ goods as a matter of preference would count as deprived.
Niemietz then suggests going back to basics by applying some elements of Rowntree’s BSA to a consensual MD measure (CBSA), or less wonkishly applying some common sense and data to the validity of expressed preferences. So for example deprivation of access to a washing machine should depend on the price of available cheap machines not the average price.
The CBSA poverty line could vary by regions and types of household (e.g. families, pensioners, students) according to local prices. Households below the expenditure line that allows you to purchase the consensus view of necessity would be classed as poor. Those above that choose not to buy those goods (they may for example spend more on clothes and cigarettes) would not.
The policy implications of a CBSA would be to more accurately draw attention to government policies that make life tougher for the poor such as planning restrictions, producer subsidies, and minimum prices. It would highlight the importance of economic growth, open markets, removing poverty traps, lowering marginal tax rates, simplifying the benefit system, removing couple penalties, school choice, workfare, and the interactions between such measures.
The rest of the monograph includes an interesting history and analysis of measures and policies in the UK.
All in all this in important and interesting contribution well worth reading to understand what is often an impenetrable debate.
*When team Cameron propose this, you heard it here first