What percentage of GDP should we be spending on defence?

This primitive form of budgeting is essentially targeting the wrong measurement because targeting the right measurement involves too much effort. Even so, differences in counting methods are going to make comparisons between countries unreliable. Not only will some countries classify a given item as defence expenditure, or as internal security, industrial support, technological research expenditure, or whatever according to taste or tradition, but some countries count GDP by output and others by expenditure. The latter inconsistency makes a big difference if governments pay for work that is not done, as happened during the pandemic. So people who advocate such targeting seem to believe that if we express an unreliable figure as a percentage of another unreliable figure, it somehow becomes meaningful at the margin when we are deciding whether a country has spent 1.9% or 2% of its GDP on defence.

Even if it were easy to calculate defence spending as a proportion of national income; what is difficult is to calculate the effectiveness of the spending. I seem to recall that The UK Ministry of Defence employs more administrators than armed forces. Meanwhile it has for decades had absurdly profligate procurement methods. There seem to be innumerable departments, all of which are entitled to demand that a particular piece of kit should be modified in this or that manner, in case of that or this eventuality, so that by the time the kit is actually delivered, several years late and many millions over budget, it has become an expensive jack of all trades, master of none, when it could have been on time, on budget and a master of one. Like the NHS, defence spending has become a sponge that soaks up all the money that is thrown at it whilst simultaneously contriving to produce less and less effective results.

Tactical defence today requires a lot of cheap kit that works all the time, not a little expensive kit that works some of the time when its hugely complicated folderols can be bothered. Consider the impact of drones upon modern warfare. Take a piece of kit that costs next to nothing and blow up something hideously expensive. Take a piece of kit that you can make in a week and blow up something that took several years to build. Figure out how you can get more ammunition when you fired off your entire reserves or sent them as aid to Ukraine. Take a little time to work out why migrants don’t find empty army barracks to be acceptable accommodation (after you’ve figured out why the barracks were empty in the first place). Find some way of noticing that if you can afford an aircraft carrier you need to be able to afford aircraft to sail on it, because it’s not much use without.

And, having done all that, you will be in a better state than if you had met a nominal percentage target for expenditure.