Looking back on my time in the academia of economics at university, one of the things that strikes me is the relative lack of examination of data – amongst other things. Textbooks are full of graphs – supply and demand, purchasing power parity, unemployment against inflation, inflation against interest rates, IS-LM, you name it. But inspite of all of these graphs in the books, I can’t recall ever seeing the data behind these graphs.
Intuitively, things can seem to make sense when explained through. One of the biggest assumptions that is used in economics is ceteris paribus – ‘all other things being equal’. The murkier the waters become in mainstream economics, the more one realises that actually there are many things that are anything but equal, and thus the flaws in the models emerge. It’s also that area where political biases are spectacularly exposed – the biggest example being around immigration. Free movement of capital, labour and knowledge? What business is it of the state regarding who a firm chooses to employ or not employ? Yet the blunt application of such principles give the feel of such economists living in a social vacuum – the firm being able to hire and fire at will, not taking into account the human impact of the instability & the worries that such conditions have on people’s health. I can’t help but feel that the mindset of “firms externalising as many of their costs, paying their workers as little as possible thus leading to maximised profits = good” was taught. Little consideration was given to the impact that this would have on how the workforce performed. If you pay people more and treat them better, won’t there be a greater incentive for them to be more productive? Won’t that also increase the calibre of people who choose to work for you, putting you ahead of the competition? Who are the business leaders who will stick their necks out on the line, say they want to pay more than the market rate in order to attract higher calibre staff, deliver better products and services & more satisfied customers?
This in part comes back down to the title of this post: What does the data say? When I started working part time during my college days in the mid-1990s, the local supermarket had just moved off from manual pricing – i.e. where every single product needed a little sticky label on it for the checkout staff to punch in the price. One of my local newsagents still does this today. The local supermarket has since moved on to a system of scanning everything, enabling things like automated stock control. Amongst other things, this allows firms to look at shopping trends for individual products and potentially compare them against a whole host of different factors – such as an advertising campaign by a manufacturer, a discounting campaign by a local competitor to things like the weather or the success/failure for the England football team to qualify for a given tournament.
This got me thinking back to the graphs I was talking about at the top. What would all the demand and supply graphs look like if all – ALL of that automated data was plotted on axes? Which products would demonstrate that if you raised the prices, demand would fall? Which products would end up doing the opposite? What would the elasticities look like – i.e. which are the ones that would maintain a relatively constant level of demand irrespective of the price changes? Which are the ones that would change significantly at even the smallest of price changes? Would the graphs be different for big retails vs small retailers selling the same products? What about comparing shop retail versus internet retail? Are there any differences there?
Part of the thinking behind this is a result of hanging around with scientists – the ones that study the traditional ones rather than the social ones. Amongst other things, the sceptical mindset & the desire to test stuff is one of the things I’ve come away with – for example throwaway comments such as “That’s about as useful as a chocolate teapot!” Why would a chocolate teapot be useless? Because when you put hot water into it, it would melt…wouldn’t it? Wouldn’t it?!? The only way to find out is to make one! (See the Naked Scientists’ full article on this here).
Something like this I think would be a fascinating study. The problems are that this would be a stupendously labour-intensive study collecting the data in the first place (because not everywhere may be collecting it and those that may be, may not be collecting it in a form that is easily useable), and secondly, firms may not be willing to give such data away (due to commercial considerations) or may want to charge for it. But the technology is there to collect, process and analyse the data and break it down/disaggregate it for a whole host of different things. Making those data sets accessible I hope would allow researchers to test and retest some of those assumptions that mainstream economists take for granted. What would the analysis of all of this data tell us? Would the data reinforce what mainstream economics already teaches or would it pull the rug entirely from under it?