The LSE has a lot to answer for

Share this article

What have Rachel Reeeves, Yvette Cooper and Ed Miliband got in common? Yes, they are experienced politicians, with a combined total of 66 unbroken years in the House of Commons. Yes, they occupy some of the key posts in Keir Starmer’s Government and will probably still be Cabinet members after a change of leadership.

I wasn’t thinking on those lines. What interests me is that these three all successfully completed the MSc Economics programme at the London School of Economics, but are prepared to sign up to policies which seem to fly in the face of economic reasoning.

The LSE programme is amongst the most demanding in the world. Although it will obviously have changed a bit over the years, its basic structure remains what it was 30 years or more ago.

There is macroeconomics, which in the early 1990s (when Miliband and Cooper were chewing their biros in the examination hall), was some blend of New Classical and Real Business Cycle theory. In the 2000s, when Reeves was at LSE, it would have been Dynamic Stochastic General Equilibrium – highly abstract mathematical modelling of the whole economy moving through time. In addition, LSE students would have covered chunks of microeconomics, probably game theory and a large dose of welfare economics (much of which tells us that an economy needs to meet ludicrously impossible conditions to perform at its best, and thus suffers from a long litany of ‘market failures’ which we need government to correct). Finally, they would have spent a lot of time learning and practising econometrics – statistical methods applied to economic data to test economic propositions and measure the responsiveness of one variable to changes in others. This area has a big appeal to many economics nerds, as it seems to offer a scientific approach inaccessible to outsiders. However, like other social sciences where there are only limited opportunities for controlled experiments, it has significant problems of reproducibility and replicability, and depends so heavily on the quality of data, the assumptions made and the significance tests applied that there is no real consensus on many empirical issues.

Many bright people pass successfully through programmes like the LSE masters. Yet they often seem to lack what you might call economic common sense.

There are some fairly basic ideas in economics which you can trace back to Adam Smith and classical liberalism:

Prices act as a signal; raising the price of something will tend to increase its supply, but reduce the demand for it. When governments suppress this signal through artificially raising or lowering a price, this almost always brings problems in its wake.

Profits are another important signal: when profits are high, it suggests further investment in an area is worthwhile, either by incumbent firms or – even more importantly – by new entrants. When governments either reduce profits artificially through taxes and controls, or raise them through subsidies and protection, markets will misallocate resources and thus reduce productivity and economic growth.

Voluntary exchange, whether within a country or between countries, usually makes both parties better off. Thus restrictions on international trade, and limitations on free choice within a country, are bad for us.

There are always trade-offs because every economic choice has an opportunity cost – what you give up when you choose one thing rather than another. As Milton Friedman put it, there’s no such thing as a free lunch.

Flowing from this, that there are always unintended consequences of decisions made by governments, businesses and individuals – because they alter incentives and thus behaviour in ways which nobody can foresee with accuracy.

The optimal level of some bad thing – let’s say crime – is never zero, as the resources required to bring it down further from a low level increase exponentially as crime falls. It’s opportunity cost again.

You know, stuff like this. I could go on. But my point is that these principles have somehow got lost in the way advanced training in economics has developed. Cooper, Miliband and Reeves are bright enough people in the academic sense  – they jumped through all the hoops, which many couldn’t –  but somehow don’t seem to have internalised the fundamental lessons of economics. Many of the specialist civil servants offering economic advice to ministers will have undergone similar training, at LSE, Warwick or UCL. They are all apparently signed up to policies which are almost guaranteed to produce bad results. For example:

Raising minimum wages sharply for young people and increasing employer national insurance contributions will inevitably make employing them less attractive. Hence unemployment and inactivity.

Restricting the type of contracts landlords can offer and limiting their power to set rents will exacerbate the shortage of rented housing.

Trying to persuade supermarkets to hold down prices of key goods (in Scotland, the SNP is trying to do this with the force of law, which Westminster could forbid) to ‘fight inflation’ didn’t work in Venezuela and won’t work here. It will have consequences such as reductions in quality of targeted goods or increases in prices on other goods.

Windfall taxes on profits and restrictions on oil extraction and fracking will increase our dependence on imported oil from regimes we purport to oppose.

Nationalising the railways and our remaining steel industry is unlikely to make them more efficient and will be a continuing drain on the taxpayer.

Above all, pursuing the impossible dream of Net Zero will keep energy prices so high that what remains of manufacturing and much of agriculture will find it impossible to compete in world and domestic markets. And it will do nothing to slow down global climate change.

I see that the LSE is charging £41,000 for its MSc in 2026-27. Successful completion will hopefully open the door to a range of lucrative professional careers, and justify the cost. But will it increase its graduates’ understanding of what policies we ought to be following? I’m not so sure.

Share this article