Stay the course on the NPS
Business Standard, 3 December 2018
India has very few civil servants, but they are generously paid. The traditional civil servants pension was extremely expensive. In the late 1990s, these problems were understood, which led up to the NPS reform. The NPS was a fiscally expensive solution for one generation. There was a loss of continuity in the reforms, and the armed forces drifted into one rank one pension (OROP). Now we see some complaints by young civil servants who do not like the scheme. Every political leadership from 2002 onwards has fought for the NPS. We must stay the course.
The Indian State is a rather small one in terms of recruitment. However, the wage structure is one with sub-market wages at the top and high wages at the bottom. As an example, private schools hire teachers at lower prices than government schools. As there are very few employees at the top, the overall expenditure per worker is much higher than private sector benchmarks.
The traditional civil servants pension was a defined benefit at about half of the wage at retirement. In the 1990s, there was an explosive trajectory of sharp growth in pension expenditures. Particularly with the armed forces and the railways, pension payments were growing much faster than wage payments.
A promise to pay a monthly pension is akin to a government bond with monthly coupons. A key milestone in this journey was the estimation of the `implicit pension debt'. The Ministry of Finance and ADB funded a household survey through which the number of civil servants and pensioners was estimated. This was used in a paper by Surendra Dave and Gautam Bhardwaj, which estimated that the implicit pension debt was about 65 per cent of GDP.
In 1999, the Ministry of Social Justice created Project OASIS, led by Surendra Dave. This designed the New Pension System (NPS). The NPS was a package deal with a 10% wage hike to civil servants and a shift into a clean defined-contribution system. The staff work that led up to the NPS included discussion with the upper echelons of the navy and the air force, who were enthusiastic about it, but the project planning involved doing this for the armed forces after the main system stabilised. The reform was championed by Yashwant Sinha, Jaswant Singh, L. K. Advani, and approved by A. B. Vajpayee on 12 December 2002. All recruits of the government from 1/1/2004 were to be placed into the NPS.
When the UPA took charge in 2004, all these questions were re-evaluated and re-endorsed. The CPI-M was part of the coalition and was not keen on the reform, and the new legislation could not be passed. Hence, the implementation began through a sound set of legal instruments and contracts with service providers. It was only in 2013 that the law was passed, and PFRDA became a statutory regulator of the service providers.
Unlike many pension reforms elsewhere in the world, there was no decline in pension payments to existing workers or pensioners. Hence, NPS was a fiscally expensive reform. For one generation, the government is paying contributions to new workers (with a 10% wage hike) and pensions to those hired earlier. The fiscal gains only arises from deaths of employees hired prior to 1/1/2004. These gains would be spread over the 75 years starting from 2004.
In this period, the Indian reforms process looked great. It was a story of empirical evidence, fundamental change, deep administrative transformation, and engagement with the entire political leadership across two rival political parties. This showcased India's ability to confront and solve difficult problems. Things look less good in the following period.
In the early days of the NPS reform, the armed forces were always part of the plan. As emphasised earlier, discussions had taken place with the navy and the air force, and they were on board. The idea was that this implementation would be done after the institutional structures were working for civil servants. However, there was a loss of institutional memory in the Ministry of Finance and this was not carried through.
Demands for `one rank one pension' erupted late in UPA-2. Both parties acquiesced, without making any calculations. In 2015, Renuka Sane and I estimated that this decision involved very large costs. Not only was the reform not done for one component (armed forces), things were actually made much worse for this component. OROP was sold in jingoistic fervour, but it has actually weakened the fiscal capacity and thus the military strength of the Indian State. We have traded off borrowing capacity in a future conflict in return for better paid employees.
These problems fall in the context of the larger question of off-balance-sheet liabilities of the Indian State. In addition to the visible stock of bonds issued by the government, there is a large stock of promises that various elements of government have made. We understand and worry about overt debt and bond issuance. We should be even more worried about promises.
It is the prerogative of the political leadership to make and modify the promises of the Indian State. But these decisions should be made under conditions of sound information. In the NPS story, the leadership of both parties absorbed the statistical analysis of multiple policy alternatives, before approving the reform with a 10% wage hike and a 20% contribution rate. In many other situations, the Indian policy process has rushed to decisions without calculations. There is much to be said for paralysis through analysis. We should insist on seeing net present value estimates over 75 year horizons.
How can the fiscal system induce greater thought about off-balance-sheet liabilities? A bond market with voluntary buyers, and the Public Debt Management Agency (PDMA), are required. Voluntary buyers of bonds will worry about fiscal stress deep into the future. The PDMA will engage with buyers of bonds and will bring the bond market perspective into the policy process. This will generate checks and balances to bear upon policy decisions.
Back up to Ajay Shah's 2018 media page
Back up to Ajay Shah's home page