Thursday, July 27, 2017

Defence expenditure: Fantasy replaces deft planning

This article was published in the DNA, dated 27th of Jul, 2017

In the recent Unified Commanders’ Conference, the Indian military has reportedly asked for an allocation of Rs 26.84 lakh crore for the next 5 years. To put the figure in context, this represents a number that is almost triple of the previous five years’ allocation. The annual simple average of the demand made this time (Rs 26.84 lakh crore over 5 years) would be double the defence budget allocated for the current fiscal (Rs 2.74 lakh crore). The numbers, analysed whichever way, are staggering.
Let us look at the headline numbers. At current levels, defence expenditure constitutes 2.1 per cent of the GDP. Assuming a 10 per cent nominal growth in GDP (6-7 per cent real GDP growth plus 3-4 per cent deflator, or inflation), the demand for 26.84 lakh crore represents 3.3 per cent of total GDP generated over the next five years. Realistically though, expenditure cannot go up overnight from 2.1 per cent of the GDP this year to 3.3 per cent next year. They will go up gradually, with a “hockey stick” (exponential) increase towards the latter half. However, that also means, given the same numbers (expenditure demand and GDP, over five years), the ratio will likely touch 5-6 per cent in 2022. Very few countries that could manage to sustain such levels of 3 per cent of GDP over 4-5 years in the late 80s ended up with a bankrupt treasury in the 90s.
Next, how do these numbers stack up on a global comparative basis? It’s an old chestnut in military policymaking, ie, India doesn’t spend enough on defence. Is that correct? There are two benchmarks typically used.
First, defence expenditure as a percentage of GDP, popularly used, but largely limited in its real policymaking import, as it is not truly reflective of fiscal affordability. As per the World Bank, the numbers stack up as follows: China (1.94 per cent), East Asia-Pacific (1.73 per cent), European Union (1.48 per cent), Republic of Korea (2.64 per cent), Pakistan (3.55 per cent), Russia (4.86 per cent), Turkey (2.13 per cent), Israel (2.38 per cent) and US (3.30 per cent)
India’s very much in the global ballpark. We are higher than China (and Asia-Pacific average), lower than Pakistan, and in the same range as Turkey and South Korea (both having complex military challenges). Major countries that spend significantly more — US, Russia, Israel — are characterised by very large domestic Military Industrial Complexes (MIC). Military expenditure accounts for large investments and employment in these economies. The situation is quite the opposite in India.
Second, share of defence expenditure in total government expenditure — less popular, but a lot more insightful for policymaking, as it reflects fiscal constraints.
It is here that the real constraints become starkly visible. India already spends a very high proportion of government revenues on defence. In fact, defence is the single biggest item in the Union Budget, outside of debt servicing.
The only major country higher is Pakistan. In a country with significant deficits in social and economic capital, the fiscal space for a bigger share of the pie for defence simply does not exist.
Lastly, India’s high import-intensity of defence expenditure makes it an inefficient medium to channelise Keynesian boosts to the economy. With 70 per cent of all defence capital expenditure spent on imports (India’s been the top weapons importer in the world now), and low multiplier on military imports, the ability of policymakers to allocate part of government “pump priming” expenditure to defence is absent.
In a nutshell, plans with numbers like Rs 26.84 lakh crore are pipe dreams rather than effective plans. On top of that, we have a severe issue in terms of quality of military spends.
In the last 5-6 years, 45-50 per cent of the defence budget has been spent on personnel costs (salaries and pensions). This component is expected to grow exponentially, thanks to more “boots on ground” — Indian Army has expanded by 25 per cent in the last 15 years. And, now the new big daddy of Budget Expenditure is OROP. Increase in personnel costs in the years to come would far outstrip nominal GDP growth.
This is not new, or even surprising. A generalist MoD bureaucracy and a (largely) hands-off political class have let the military get away with fantastic wish lists in the name of planning. Unless structural changes are brought about in organisation and management, we will continue to have hopes as plans in India’s military.


Wednesday, July 5, 2017

Obsession with Fiscal Deficit is Hurting Growth, and hurting democracy too

Farm loan waivers have reignited concerns of fiscal deficit (FD), which is essentially the difference between the government’s revenues and its expenditure. While the focus on FD reaches its zenith just before the Union Budget, it is a core element of our economic narrative. Is 3.5 per cent too high, or 3.2 per cent? From rating agencies to economists to pundits, FD has assumed divine dimensions, with a 0.2-per cent slippage deemed blasphemous.
The question is, how important is FD to growth and welfare? Textbooks are divided. Neoclassical theories frown upon FD like plague, Keynesian theories privilege FD as a preferred policy tool, and Ricardian theories preach complete indifference to the concept itself!
The neoclassical view, the source of most punditry on the subject, base FD’s divinity on four basic canons.
One, debt (incurred to fund FD) growing faster than income (GDP) is unsustainable. Two, excessive debt leaves us vulnerable to foreign investors pulling out, triggering an economic crisis. Three, high FD causes an increase in interest rates, increasing the cost of capital for the private sector. And four, the moral argument of burdening our children to finance our consumption today.
Not trusting “irresponsible politicians” to protect the holy covenants, a Fiscal Responsibility and Budget Management (FRBM) law was enacted in 2003 to enforce discipline through prescribed targets. Unfortunately, this has meant elected representatives are constrained on a key policy tool to react to economic distress — because of technocratic limits.
Beyond the point of (democratic) principles, obsession with FD is untenable on multiple grounds. To start with, the divinity canons are tenuous in the Indian context. The Government of India (GoI) doesn’t borrow overseas (from foreign investors) to fund FD. It is funded entirely out of domestic savings. With capital account controls, GoI bonds represent the highest-quality investment for the domestic investor. Ergo, GoI can run high levels of FD without triggering a currency crisis. This is the exact opposite of oft-quoted examples — Russia, Argentina, Greece — where high FD triggered economic crises. In all these cases, governments depended heavily on foreign investors to fund FD.
When the basic issue is a lack of demand, the antidote has to stimulate demand. Private sector investment decisions are not predicated on a 1-per cent higher (or lower) interest rate, but capacity utilisation and expected future demand. Obsession with an incremental 0.5-per cent FD jeopardises growth and ability to service debt in the long term.
A hard FD target acts as what Harvard professor Lawrence Summers calls a harmful ‘repression of budget deficit’. To meet the target, the government imposes arbitrary spending cuts on maintenance and critical investments in social and physical infrastructure. The result is a lowering of national productivity, leading to lower long-term growth. We have seen this with distressing regularity in India, where FMs have cut investments to meet FD targets.
Currently, the biggest challenge confronting the economy today concerns jobs. Boosting job creation by 3-4 per cent (and nominal GDP by a bit more) will do a lot more to balance the FD in the future (after all, FD’s real relevance is as a percentage of GDP) than targeting a sacrosanct number today.
Above all, causality between low FD and growth is extremely tenuous in terms of data in India. High growth has typically been a result of higher demand creation. Without supporting data, to base policy on ideology is simply bad policy-making.
The bigger philosophical question though is the usurping of sovereign democratic powers by an unelected technocratic elite. The FRBM Act was a primer. Recently, the NK Singh Committee appointed to recommend improvements to FRBM has suggested setting up of an autonomous Fiscal Council (staffed by technocrats) to administer and monitor FD targets set out in the law. This would represent a further erosion of political authority to tackle a political economy question.

Forex Reserves to the rescue

The Public Sector Bank (PSB) recapitalisation debate continues unabated. Clearly, “simple” options — amalgamation, Bad Bank, Indradhanush — are “dead on arrival”. It is also quite clear that the government has limited appetite to infuse fresh capital, given fiscal constraints. Depressed valuations make raising capital from public markets a challenge, especially for weaker PSBs. So, is there a game left in town?
There could be, via direct Central Bank (CB) intervention. In recent times, both in the US and in Europe, CBs have bailed out banks, primarily through purchases of stressed assets (mortgage-backed securities in US, Greek bonds in Europe). The modus operandi is simple. Say, a bank needs to write off $100 of stressed assets, and it does not have the capital to do so. The CB prints $100 of fresh money uses it to buy out the stressed assets, enabling the bank to remain solvent. This, of course, means an increase in CB balance sheet and consequently money supply in the economy, which could fuel inflation. But in the low inflation scenarios in US/Europe for many years, it has been a politically acceptable risk to take.
Can RBI do this in India? Technically yes, practically difficult. India, unlike US/Europe, typically suffers from high inflation. While inflation has been tapering off recently, it is often vulnerable to systemic shocks like oil prices and bad monsoons. As recently as the last monetary policy, RBI flagged off upside risks on inflation. Ergo, RBI printing more money to buy back NPAs (non-performing assets) from banks has significant inflation risks, making it politically untenable (as well as running counter to RBI’s Inflation Targeting mandate).
There could, however, be another way, and that is to use RBI’s foreign exchange reserves to directly recapitalise banks. How does this work? Today, RBI holds $380 billion of Fx reserves as assets in its balance sheet. These are typically deployed in a variety of financial market instruments — predominantly hi-grade Treasuries (like US), Bonds and Gold. RBI could use a part of these reserves to buy capital instruments issued by PSBs. This would achieve a couple of objectives:
  • It doesn’t expand RBI’s balance sheet/money supply — it simply swaps one type of securities (say, US Treasuries) for another (Indian Bank Capital, or IBC). There is no increase in money supply and downstream pressure on inflation.
  • It keeps the funding off the government’s budget. Consequently, it does not deteriorate fiscal ratios. Let’s look at the broad math:
Total Banking NPA — estimated at $180 billion
Haircut required to make the loans viable (say) — 50 per cent
Capital required for write-offs — $90 billion
Now, RBI uses (say) $45 billion of its reserves to buy IBC, providing 50 per cent of the required capital. Such a large capital injection will lift underlying sentiments and stock prices, enabling PSBs to tap the capital markets for the balance $45 billion. Coupled with other reforms (Bankruptcy Code, RBI cover for bankers to settle NPAs etc), PSBs could start on a cleaner slate. In a few years, stock prices could rise sufficiently for RBI to sell off the IBC at a profit, similar to what happened with the TARP programme (to bail out banks in 2008) in the US.
There is no free lunch, and this approach isn’t without risks. The biggest is the fact that Fx reserves are not sovereign wealth. RBI has liabilities (primarily remittance claims from foreign investors/importers) against these reserves. These reserves are usually deployed in liquid, high-quality securities to enable RBI to have liquidity in times of stress. As IBCs would be illiquid, Private Equity-type investments, it would mean RBI has proportionately fewer reserves to call upon in an emergency.
That being said, $45 billion constitutes a fairly small part (12 per cent) of India’s reserves, and can be relatively quickly replenished given the strong capital flows that we are seeing today. A repaired banking system will add to the “India story” and further enhance those flows.
This isn’t an idea that’s new, but one whose time certainly has come (to at least being evaluated). India’s PSBs need capital, and to use a Sherlockian adage, once all wickers have been burnt, the one left standing is the most plausible one. Using Fx reserves is a plausible wicker left to recapitalise India’s PSBs.

Wednesday, June 7, 2017

Can central bank stem NPA rot?

This was published in the DNA on the 29th of May 2017

Indian banking is often in the news — sometimes as entertainment (around high-profile failed airline promoters), but mostly serious. The government’s recent NPA Resolution Amendment to the Banking Regulation Act is a serious development, which comes not a day too soon. NPAs now, at nearly Rs 9.6 lakh crore, account for 15 per cent of all bank loans and constitute a serious bottleneck to growth. However, is it kosher for RBI, as does the Central Bank (CB), to get operationally involved in the resolution of NPAs? And, is it enough?
The short answer to the first question is, absolutely. The puritan argument is that it represents a conflict of interest for the CB. A regulator cannot be seen to be resolving what is essentially a commercial decision of a regulated entity. However, taken to its natural conclusion, this approach will result in the banking industry proving the old chestnut “operation successful, patient died”. Ergo, CBs often get involved. The Bank of Japan (BoJ) recapitaliSing Japanese banks in the late 1990’s, US Fed acting as a lead arranger of rescue finance for LTCM in 1998 and the US Fed (2008) buyback of $1.5 trillion of distressed mortgage securities — these are but a few recent instances of active CB involvement in bank bailouts. In short, this is par for the course.
Now, for the second (and real) question — is it enough? A basic background first.
What are the basic steps involved in resolving a bad loan (NPA) issue? It normally consists of three basic rules:
Firstly, recognising the bad loan. Secondly, estimating and affecting a haircut on the loan, either to make the underlying business viable or to sell off the loan, and thirdly, generating capital to absorb the haircut.
Over the last few years, RBI has done stellar work in recognising NPAs. However, it is in the resolution of these loans where the ball got stuck.
First, thanks to the scourge of the 3 Cs (CBI, CAG, CVC), PSB bankers have been loathe to structuring haircuts for any loan gone bad. Arrests of senior bankers (including an ex-chairman of IDBI Bank) have further queered the pitch. Second, a limited tenure of most PSB bankers have meant that they have found it easier to simply evergreen the loans. In simple parlance, leave the problem for their successors. As a result, a stressed loan of (say) 100, after unpaid interest, interest on interest, fines on delays, balloons up to 200 or more by the time someone starts taking action.
This is where the new Ordinance helps. With this, bankers have received RBI cover to take decisions. RBI has a high (and deserved) reputation of integrity and professionalism. Haircuts agreed to by bankers under the direction, supervision and operational monitoring of RBI are unlikely to invite allegations of personal corruption that they could otherwise. While the process is still complex, with relatively few cases accounting for the bulk of bad loans (the top 20 cases account for over half of stressed loans by value), decisions should now be happening at a faster clip.
This brings us to the last, and by no means the least, of the issues — capital. Once the bank has identified the bad loan and determined a fair haircut value for the asset — it needs capital to recognise the haircut on its balance sheet.
Unfortunately, bank recapitalisation has progressed inadequately. The total outlay in the Union Budget for bank recapitalisation is Rs 10,000 crore — perhaps 20 per cent of what the banks really need. Cute acronyms (like Indradhanush) and toothless governance structures (like Bank Boards Bureau) have been tried. If that seems Tughlakian, policymakers are talking about a bigger one — consolidation. The merger of associate banks with SBI is being touted as a precursor. Easy in theory, but has a rather simple tautological fallacy — the banking system simply does not have surplus capital to absorb the aggregate PSB capital gap. It’s a classic case of adding 2 and 2, repeatedly, hoping each time that the answer somehow will be 10! SBI quarterly results showed it up starkly — nearly Rs 3000 crore of standalone profits, but Rs 3000 crore of loss on a consolidated (with the subsidiaries) basis. In short, even a small amalgamation pulled the giant down. Clearly, consolidation is not a likely solution.
Unfortunately, the new ordinance does nothing on capital. The solution has to be about someone writing out a big cheque, which only the government can plausibly do so.
The new Ordinance covers two of the three steps required for a resolution. It’s a big, necessary step forward. But it is not sufficient to resolve the issue — till the government recapitalises PSBs.

Friday, May 26, 2017

Human Shield - yet another necessary evil in India's fight against terrorism

It doesnt take much to generate outrage these days, thanks to both mainstream and social media. However, the kerfuffle over the "human shield" incident has been a topic worthy of debate. Using a civilian as a shield to facilitate tricky counter insurgency operations is neither new nor unusual - either in India or outside - but the images of a civilian tied to the bonnet of the jeep driving through villages is undeniably evokes strong reactions.

Besides the usual battle lines between "liberals" and "right wing", an interesting divergence has been in the military circles, with several ex-servicemen coming out against the modus operandi adopted by Maj Gogoi in this case. They have quoted Geneva Convention, Indian Army ethos, counter insurgency doctrines and more. As Lt Gen H S Panag, ex-Army Commander, summarised,  
"The image of an alleged ‘stone pelter’ tied in front of a vehicle as a ‘human shield’, will forever haunt the Indian Army and the nation"
Honour is essential to a professional army. But the top brass in Indian defence, at both the military and the ministry level, seem to have forgotten that.
But is it really so? Is the human shield the first time a somewhat distasteful tactic has been used in countering insurgency? Or is it but another in a series of such innovations that the Indian security establishment has had to adopt to counter the threat of insurgencies and terror? Lets look at a few examples.


The Mizo insurgency was one of the first such challenges faced by independent India. On Mar 2, 1966, the Mizo National Army (MNA), the primary Mizo insurgency group, overran large parts of Aizawl, the capital of Mizoram, as well as a few smaller towns around it. It included the Treasury, Armoury as well as the Assam Rifles headquarters. After the initial attempts by the Indian Army to relieve the city was beaten back, the Prime Minister Indira Gandhi ordered the Indian Air Force (IAF) to conduct operations. In the first (and till now, only) case, IAF fighters strafed Aizawl with guns and bombs, destroyed large parts of the city, and killed several civilians. But the operation succeeded, MNA melted away into the jungles, and the might of the Indian state was re-established in Aizawl.

In 1967, the Eastern Command of the Army, led by Gen Maneckshaw, launched a programme called "regrouping of villages". It basically meant driving out villagers from villages and grouping them in fewer hamlets (called Progressive and Protected Villages, or PPV). The idea was it would be easier to monitor fewer number of villages and therefore cut off the support to MNA from the populace living in the rural hinterland. some 500 villages were regrouped into 100 odd PPVs - it accounted for 95% of Aizawl's rural population. Did it work? Open question, but the last PPV was dismantled only in 1980, 6 years before the Mizo Accord that ended the insurgency. It was one of a large bouquet of measures required to tackle a well armed insurgent group.


One of the deadliest challenges faced by the Indian state. KPS Gill is widely credited for ending the insurgency, using unconventional means that would not pass muster in "normal" situations. One of them was the "hostage" strategy. A tactic that was a straight lift from the terrorist playbook, it involved picking up (either physically by the police, or through more discrete means) members of the families of known terrorists. Keeping such hostages achieved two purposes. One, it meant that terrorists left the families of policemen (most of whom came from rural Punjab) alone. Two, it worked as a lever for select influential leaders of Khalistan groups to come forward and assist the police in exchange for the security of their families. This was carried out widely in rural Punjab, which is where bulk of the Khalistan leadership emanated from.
The result was a resounding success. This, along with many other measures, helped end the militancy by 1995, barely 11 years after Op Bluestar. In 1985, it was the greatest national security threat faced by India, and no one could imagine that it would be defeated by 1995.

Naxalite movement in West Bengal

There was a time in the late '60/early '70s when the city of Calcutta was literally hostage to the depradations of Naxalites. Policemen were killed in broad daylight, businessmen killed and driven off and the slogan of "China's Chairman is our Chairman" adorned most walls of the city. Biggest issue was the urban middle class sympathy for the movement - some of the brightest young boys from middle/upper middle class families provided leadership to the Naxals. Under an unconventional police chief, Ranjit Gupta, the state (and central) govt turned the tide using what can be mildly termed strong arm tactics. To start with, the police started randomly picking up students from hostels known for their Left wing activism. Many of them were sent to jail and kept without trial for years. Several were simply shot in cold blood. Many families were given a stark choice - to send their boy out of Bengal (preferably out of India), or see him incarcerated or killed soon. Most families exercised the former option. It was a campaign that sent shivers down the cities of Bengal, but it slowly but surely cut the oxygen of fresh recruits to the Naxal movement. By the mid/late '70s, Naxalism had ceased to be a threat to the state.

These are but a few instances. Counter Insurgency (CI) is a complex, dirty business. Hearts and minds is identified as key to CI, but any such campaign can only succeed once the majesty of the state has been established. A neutral populace will tend to side with the group wielding the most coercive power.

Ergo, its perhaps unfair to judge the officer who took the Human Shield call. The Indian Army (and state) have adopted far more distasteful tactics in defence of the state in the past, and succeeded. They were necessary to defend the freedoms and ideas of India for the vast majority of India's citizens.

Sunday, May 21, 2017

Strategic Partnership model for Defence - Modi govt's SEZ moment?

After a long and tortuous journey, the Ministry of Defence (MoD) recently finalised the Strategic Partnership (SP) model for defence manufacturing. The salient points of the new policy are as follows:

1. Four key systems - Single engine fighter aircraft, helicopters, submarines and Main Battle Tank/Armoured Fighting Vehicles - have been opened up for private sector manufacturing. While the first two have been opened up exclusively for private sector manufacturing, the last two are open for PSU/OFB to bid for as well.

2. There will be ONE SP per project selected.

3. The SP will require tie-ups with foreign Original Equipment Manufacturers (OEM), to cover manufacturing, transfer of technology (ToT), assistance in training skilled human resources and other support.

The putative objectives of the policy are clear and have been discussed for years. Create a Military Industrial Complex (MIC) in India, provide competition to an inefficient PSU eco-system, and prmote indigenous solutions to our large defence requirements. The question is, does the policy do so?

Shorn of the hype and hyperbole, not only does the policy not seem to achieve any of the stated objectives, it in fact lends itself to massive rent-seeking opportunities (or at least motivations for the same). At its core, the policy seems geared to simply replace a public sector license-manufacturing monopoly into a private sector license manufacturing monopoly. 

Lets look at the structural elements. 

First, the defence industry, by its very nature, is a monopsony (a market that has only one buyer, and multiple sellers). Typical other examples of monopsony markets are Railway engines, British National Health Service. In such a market, the buyer can maximise value from this market by promoting a high level of competition among many sellers. The issue though is, military technologies, by their very nature, are available with few companies/entities. Ergo, in the best of times, it is difficult to find enough competing sellers to foster genuine competition. Governments around the world therefore try various incentives to foster competition. The new SP policy however, reverses the logic, by creating monopolies in key areas. In short, now the monopsony buyer (the government) creates a monopoly seller to buy from, instead of fostering wide competition to derive maximum value for itself!

Second, the structure envisaged in the new policy is exactly what has prevailed for many decades in India. Typically, the government (and the military) ask a bunch of foreign OEMs to bid for a contract, select one from those who respond (or remain left in the fray after many years of dilly dallying), and then designate a PSU (or OFB) to license produce the product in India. From Mig21s to Su30s, T55s to T90s - this has been the model adopted. The new policy simply replaces the PSU license producer with a selected PRivate Sector SP. The process remains the same - a foreign OEM provides the know-how, a partner Indian SP sets up the manufacturing facilities to license produce the system in India. In other words, convert a public sector monopoly into a private sector monopoly!

Third, the policy supposes that somehow private ownership is the panacea of all ills plaguing the defence industry. Problem is, private sector operates on principles of profit maximisation (as it should). When a policy seems geared to award monopoly status around license manufacturing of a major system owned by a foreign OEM, there is little incentive for the SP to invest in large R&D and achieve true self sufficiency. Long years of license production of Mig21 didnt enable HAL to design a new fighter (or fighter engine) on its own, because it knew when the time comes, it will license manufacture the next generation (which it did, with the Su30). Similarly, a monopoly private company license producing the F16 (say) has no incentive to set up capabilities to design and manufacture the next generation of single engine fighter. It would rather focus on maximising profits for the assembly line by "working the system".

Many years ago, the UPA government unveiled an ambitious SEZ policy, to emulate China's success with SEZ. But thanks to the design of the same, the policy became a tool for tax avoidance and real estate - related land grabbing. Very quickly, the entire process got embroiled in litigations, allegations of corruption and controversies. It was a design defect with the policy itself. The defence SP policy portends to a similar fate

Monday, May 8, 2017

India's defence expenditure - its a problem of quality, not quantity

Do we spend enough on defence? The general narrative in the commentariat forever is a resounding NO. The point received some recent impetus when the Army Chief, General Bipin Rawat said that the military is not getting its due share of national resources. For good measure, the Chief also said that India should take lessons from China in this regard.

Lack of enough spends on defence is an old chestnut - question is, how valid or even accurate is it?

Lets look at the headline number everyone quotes (often inaccurately), military spending as a % of GDP.

Source: World Bank

As can be seen, not only is India well in line with rest of the world in terms of how much of our GDP we spend on defence, we are in fact ahead of China in this regard. Of course, given the disparity in the relative economic sizes, China spends a lot more on absolute terms, but the solution to that is to expand our own resource base (GDP) to afford a greater spend. However, do we spend too little? Data doesnt suggest that at all, in fact its quite to the contrary.

However, its not just about GDP, but also budget expenditure-intensity of defence. Finally, defence can be funded only out of the public exchequer. And there are competing demands for the limited tax-kitty that funds defence too. And it is here that India is an absolute outlier, on the negative kind.

Chinese budget numbers are not available, but India spends clearly a very high proportion of its Central Budget on defence. Comparable to the US, and only marginally short of Pakistan, where the military is in charge of government and tends to have outsized share of resources and influence.

Put the two together - proportion of GDP and proportion of central govt expenditure, and the oft-repeated lament of low spend on military doesnt stand up to scrutiny. So, what really is the issue?

In simple terms, its quality not quantity. We spend enough, but we underspend on quality. It has 2 facets.

One, we spend too much on salaries and pension. In the last 4-5 years, 45-50% of the defence budget has been spent on personnel costs. As a comparative benchmark, US spends around 25-30% of its defence budget on personnel costs. Bad news doesnt end there though. Thanks to OROP and a steady increase in the number of personnel (Indian Army has expanded around 25% in the last 15 years, a period when almost all major militaries have downsized on personnel), the pressure of salaries/pension is only going to increase.

Two, India is unique in terms of its import-intensity of expenditure on military. For years, we have been amongst the top 3/4 weapons importers in the world. Of late, we have acquired the dubious distinction of being the numero uno!

Source: SIPRI

Countries which spend more than India on defence (as proportion of GDP) - Russia, US, Israel - typically are large net exporters. The military industrial complex is a large part of their domestic economies/employment. For India, conversely, large chunks of the defence budget go to financing imports, with no network externalities in the domestic economy. Its a double whammy, where we spend a pretty large sum of money, while not creating long term benefits for the economy. There is adverse military impact too, as imported weapons are often not available when they are most needed (ammunition for Bofors gun during Kargil as a case in point).

In nutshell, Gen Rawat (and the commentariat) is completely wrong. India doesnt lag in spending on defence - we spend as much as we can afford (and some more). The issue is on quality. Unless we improve upon that, we will continue to derive sub-optimal outcomes for the considerable sums that we are spending on securing India.