Read on Twitter
Thread on my thoughts why data derived analysis on military expenditure is problematic, bought about after reading this https://upr.lse.ac.uk/articles/abstract/10.0000/cja5553/
The starting point for most analysis on regional or country military spending is normally the SIPRI datebase on military expenditure and budget though there are also others but the problem is that defence budgets encompasses a wide range of portfolio spending under it
General perception of defence budget is that it spent on buying arms and equipment but has to be borne in mind that the defence budget pays for other things directly or indirectly related to defence..
Things that are obvious are maintenance, pay for the troops, supplies, medicine but less obvious are the things like welfare assistance, social ( military spouses program, bands, display teams, sports etc) which all adds up also but not directly translate to a defence build up
Also there are things like housing, general services and admin supplies (pens, notepads etc not free unless you get it at air or defence trade shows 
) etc too under the defence budget
) etc too under the defence budget
So to go back to it, generally most databases don't go to the nuts and bolts of actually calculating where money exactly goes to improving equipment/capability but instead calculate on overall defence budget which can be problematic e.g...
when the bulk of defence expenditure actually is paying for salaries or increased to accommodate a pay raise for the troops, coincidentally here that normally happens just around an election period
The availability of actually breaking down the defence budget depends on transparency of the country on such figures though in Malaysia, you actually can get a general idea of things as the budget document in Parliament gives a general outline plus lists the number of people
Though looking at number of people covered under Mindef budget isn't too easy as the figures released does not distinguish between military personnel and civilian staff employed
So to get back on track, an increase in defence spending does not necessarily mean something on the grand scheme of things, like China's rise, Donald Trump all over place, SCS tensions etc is prompting it but...
other factors like need to pay more to troops as need to fulfil electoral promise, cost of MRO or supplies up etc, troops got deployed more often so allowances need to be paid and so on, all of which seldom features in analysis/articles based on overall defence budget
And even things such as buying actual weapons systems can be difficult as has to be seen in context, buying a whole load of AFVs not necessarily relate to China and SCS but more to concern on land invasion by neighbours but more headline grabbing if say because China/SCS
So to go back, data driven on monetary terms can be problematic not just on defence budgets but also things like arms exports, U.S listed as leading arms exporters but U.S equipment is pricy and includes full support which adds up fiscally but Russia/China cheaper so..
How do we calculate if not on money for exports, tonnage? problematic given products same type but different weight, number and types but does exporting a million assault rifle make you a bigger arms seller when it 1/1000 of value that other guy's sale of 24 attack helos?
So you can see the problem on data driven by fiscal values, unless you can delve really into the exact details, its difficult to really say what drives a particular increase or relate it to a particular situation
So to cap off, figures and numerical data to me is only a small part of the equation in defence analysis, at end of day the people/human plus various other intangible factors all count, it's comfortable to go on data as figures are clear and quantifiable while others not
but the problem is that going on instinct or 'use the force' in your analysis is always difficult as intangible compared to data plus you can't exactly put " this is going to fail because the person in charge has always been known to be incompetent " in a research paper
So to end thread, personally I'm always inclined to look at every other factor even non-directly military related rather than just go on data, have seen some military plans derailed not because of fiscal but just because the ones in charge not like the officer who proposed it...
