Sunday, September 6, 2015

On Whether We Can Cut Military Spending by 5.75Bn

In SDP's manifesto, the SDP suggested that it intends to to cut SGD $5.75 billion from our current defence spending, or approximately 44% of our defence spending to pay for the increased budget spending on welfare and other plans.

Someone asked me if this was feasible or wise?

I am not a military expert, although I did serve 2 years of National Service as an Armour Recce Commander.

However, I am good at creating Monte Carlo simulations. Which is what I did to try to answer the question of whether Singapore can our military spending by SGD $5.75bn (40%)

Picture taken from Janes:

From Janes Defence, Singapore's National Defence budget is SGD13.1Bn in 2015.

Let's compare the expenditure across the various government ministries:

From the above Data, we can determine that the top 5 biggest spenders are

Mindef: 13.1Bn (19.2%)
MOE: 12.10Bn (17.7%)
MOT: 10.85Bn (15.9%)
MOH: 9.29Bn (13.6%)
MHA: 5.00Bn (7.3%)

In comparison, with other Ministries, Mindef has the largest share of the budget at nearly 20% of Singapore's budget.

Now let us compare with a bucket list of other countries. (Data taken from here)

From the table above, we can see the following characteristics about Singapore:

1) Singapore has the 7th highest absolute Military Budget (MB) 
2) 3rd Highest MB as % of GDP, 
3) Highest MB per capita 
4) Highest MB per land area. 

So comparatively, Singapore has a relatively high budget for it's size, output and population.

Via comparisons with other countries and other ministries, it seems to suggest that Singapore has overspent on military. However, this is not enough to create an objective opinion of whether Singapore can reduce military expenditure.

You can download my model from the link below.

In the model, we can simulate the probability of an aggressor winning over Singapore by keying in various inputs.

B2: Input Singapore's Military Spending
B3: Input Singapore's GDP
B4: Select Aggressor
B5: Input Aggressor's Military Spending
B6: For use in Monte Carlo/Optimizer Simulations. Select the number of simulations, the higher, the greater the accuracy of the simulation but the longer it will take.
B7: Indicate what is the level of collateral damage to Singapore's GDP for every Military Damage Singapore Takes.

Start Attack: This simulates 1 attack in Ranges B11 to K32. Includes Short animation involving circles and lightnings
Reset: If you messed up the numbers, click this to reset everything
Monte Carlo Simulation: Runs x number of simulations, depending what you input into cell B6. Results are in table B34:J36
Optimizer: Runs 20 Monte Carlo Simulations from 1.1x of Aggressor's spending to 3.0x of Aggressor's spending.

So I ran the above simulation against aggressor M which has a military spending of 5.4Bn a year. I ran a Monte Carlo Simulation with 100 iterations. Up to a level of spending of 1.4x more than the aggressor, there was still a possibility of losing. However, beyond 1.5x, there is a 100% probability of winning against the aggressor. However, it is worth nothing that the level of losses sustained by the country (both collateral damage and military resources) decreases with each level of spending. Hence, there is some rationale for a country to end up with 3x higher spending than the aggressor. 

However, to truly optimize the level of spending, there are many other issues to consider such as whether military spending crowds out other government expenditure/investments and whether it will lead to an arms race.

In conclusion, Singapore's military spending is very high when compared to our regional neighbours and when compared between ministries. 

In a simple model like this, we cannot conclusively decide on an "optimal" level of military spending but we are able to some extent model the effects of military spending on collateral damage and probability of losing sovereignty. When compared to our neighbours, we could possibly reduce our military spending to 1.7x of Aggressor M's spending and still maintain a significant chance of victory and a deterrent effect (assuming that we only consider Aggressor M as potential military threat.).

If we use 1.7x as the magic number, Singapore's defence budget will be USD 9.17Bn which means we can cut military spending by USD 0.33Bn or SGD 0.47Bn. As such based on a forward defence and deterrence policy, There is little room to reduce military expenditure if Singapore wants to have nearly 0% chance of failure against a military aggressor.

PS: I have assumed a direct correlation between Military Spending and a nation's defensive and offensive ability. In reality, we also need to take into consideration the efficiency of Military Spending. My perspective is that if Military Spending is less efficient than the Aggressor, then we need to take look at how to improve the efficiency, however, if Military Spending is more efficient that the Aggressor, then it creates a bigger buffer. Also, there are many more assumptions which I have taken liberally here, feel free to leave your comments below on how this model can be improved further.


If you liked this, click below for more tools:

On Whether AHPETC Overpaid it's Managing Agent?

NRIC Suffix Checker

Property Valuation Tool

When to Sell My Flat Tool


The Day The Father of Accounting Rolled in His Grave


On SG Budget Babe's: The Truth About Temasek vs Chee Soon Juan's Claims



Chee Ming said...

Hi John,

I think you should include another aggressor, I (Indonesia), into the simulation.

Given that they hope to be the military stronghold in Southeast Asia, I think Indonesia might be a bigger threat to us than Malaysia.

Chan said...

Love this man!

Shawn said...

Hi John,

Great analysis. Do you think you should also include the opportunity cost to the country of National Service? I estimate there are about 40,000 NSFs who are being paid a lower wage than they could earn if they were working. Would be interested to see how you'd calculate this.

Keep up the good work!

SpaceMonster said...

Interesting analysis, although I cannot say I really understand the assumptions behind the simulation that result in the answers. Also, I wonder: is the objective of our national defence to ensure 100% win? I thought it was just as a deterrent?

Also, what about the arms race effect? If our neighbours see that we are armed to the teeth, they may then also increase their military spending, which ups the stake for us and we have to also spend more to ensure success, and so on. Most arms races are very wasteful.

Perhaps better diplomacy and neighbourliness can pave the way to less wasteful and stressful co-existence?

John Lam said...

Hi Chee Ming,

Indonesia's Military spending is lower than Malaysia's, so in this case, there is no need to run the analysis for Indonesia. However, you can still go ahead and do so on your own by downloading and using the simulation tool.