Don’t throw away Angus Deaton with the bathwater of the Riksbank prize – he’s good!
A few days ago Angus Deaton received the ‘Nobel’ prize for economics. The good thing about this is that Angus Deaton received this price. Academic economists are not good at measuring the real world – for one thing because the very process of measurement might force economists to kill their darling concepts, like ‘the representative consumer’. Angus Deaton is good at this (at a meta level, though he does now about the gritty details of measurement) and managed to convince academic economists of the importance and necessity of the (messy, political) process of measurement (in the process shredding ‘the representative consumer’ in many, diverging pieces) and the importance of using the right statistics, instead of easy shortcuts, to answer questions like: how many poor people are with us. Deaton is pretty obsessed with the answer to this last question and he suggests it is getting ever clearer, which according is important as it gives the poor an indirect voice and some political clout – they are getting visible. Also, such data are used anyway by institutions like the Worldbank, so they better be good. But in real life, the data are, though better than ever before, still very fuzzy, which among other things gives organizations too much leeway to tweak them to their advantage… Look here for a Deaton inspired recent World Bank report about how to draw the ‘poverty line’. Look here for the 2010 Deaton presidential address about ‘price indexes, inequality and the measurement of world poverty” to the American Economic Association. Yes, these are very boring, technical papers. But look at the link on top for an inspiring lecture by Angus about what the calls ‘the political economy of data’ (mind the 2013 WEA internet conference ‘the political economy of economic metrics‘, comparable name, comparable theme). Some quotes from this lecture (my messy transcript):
The way that politics gets into the measured data is not just given, data are shaped and then data shape us … This is very tied up and very messy … There are many political struggles behind the statistics, politics is often disguised as science … a-political objectivity is often a very effective political strategy … the state may delegate political decisions to statistical agencies… price indexation is probably the most familiar example of this in which the government says that benefits will be indexed to the priced index instead of that the government has to decide what the benefits have to be… We often tell our students that there is no measurement without theory. In addition there is no measurement without politics … politics are actually necessary for good measurement otherwise the thing can unravel … the statistics which are important for who get what obviously come under especially great pressure. If they are not soundly based .. they are likely to be destroyed by the political debate We need the understand the complexity