India must first understand what it spends on public sector compensation

India may be significantly underestimating the true fiscal cost of public sector compensation, raising concerns over long-term fiscal sustainability

Public sector compensation is one of the largest and most rigid components of government expenditure. Yet across countries, the true fiscal cost of public employment is often poorly measured. India is no exception — but recent analysis suggests the underestimation is far greater than previously understood. 

This issue has become pressing in today’s global environment. Public debt levels are elevated across advanced and emerging economies, interest rates are higher than in the pre-pandemic era, and fiscal policy is increasingly expected to respond to multiple demands — from energy and food security to climate transition and industrial policy. These pressures are straining fiscal frameworks, making it harder to distinguish between temporary support and permanent spending commitments. In such an environment, accurately measuring rigid expenditures becomes essential for fiscal credibility. 

At first glance, international comparisons suggest India’s public sector compensation burden is moderate, at around 3 - 3.5 per cent of gross domestic product (GDP) — below countries such as Brazil and South Africa, and comparable to several advanced economies. But such comparisons are misleading. Reported data excludes large segments of India’s public workforce and significant components of compensation. How compensation is measured matters as much as how much is spent. International standards broadly include wages, allowances, employer contributions, and in-kind benefits across general government, including subnational entities and public institutions. India’s reporting framework diverges in three important ways. 

First, coverage is incomplete. Large segments of the public workforce — contractual staff, outsourced personnel, and scheme-based workers such as ASHA and Anganwadi workers — are typically excluded from headline salary figures. Internationally, workers performing government functions are generally included, regardless of employment status. 

Second, reporting is fragmented. Compensation is spread across multiple budget heads and institutions: Grants-in-aid to autonomous bodies, separate defence pay, distinct railway and postal pension accounts, and off-budget expenditures. In stronger public financial management systems, payroll and personnel data is increasingly integrated to provide a consolidated view. 

Third, definitions are inconsistent, particularly across states. While standardised definitions ensure comparability, Indian data varies significantly across budget documents, audit reports, and central bank databases. This undermines both inter-state comparison and aggregation at the general government level. 

When these gaps are addressed, the implications are striking. Adjusted estimates suggest India’s true public sector compensation bill is substantially higher than reported. After including missing components, compensation rises from around 2.4 per cent of GDP to around 3.8 per cent for the Union, and further to about 4.6 per cent after taking into account public sector enterprise employees. At the state level, reported salary expenditure of about 2.5 per cent of GDP rises to over 4 per cent. These are not marginal revisions — they represent a fundamental remeasurement of fiscal commitments and recognise existing obligations more accurately. 

India is not alone. In Indonesia, non-salary honoraria have been estimated at nearly a third of the compensation bill. Across advanced and developing economies alike, increasing reliance on contractual and non-permanent employment has complicated compensation measurement. This has important fiscal implications. Compensation is “sticky” expenditure: Once incurred, it is difficult to reduce. Underestimating its true size can lead to overly optimistic assessments of fiscal space, miscalibrated deficit targets, and inadequate provisioning for long-term liabilities such as pensions. In federal systems like India, these risks are amplified by variation across states and the absence of a consolidated general government view. 

The timing is critical. India is preparing for the 8th Central Pay Commission, several states are reconsidering pension systems, and the Union has introduced the Unified Pension Scheme. These decisions will shape fiscal trajectories for decades. Without accurate and comprehensive data, their implications cannot be properly assessed. What lessons does international experience offer? 

First, compensation measurement must extend beyond permanent employees to include all workers performing public functions, regardless of contractual form or institutional location. 

Second, definitions and classifications must be standardised. Aligning with international frameworks would make compensation data comparable across states, levels of government, and over time. 

Third, data systems must be integrated. Many countries are moving towards linking personnel, payroll, and treasury systems, enabling real-time tracking of workforce size and compensation costs. India’s digital public infrastructure provides an opportunity to do so at scale. 

Fourth, pension transparency must improve. International best practice increasingly includes actuarial projections of long-term pension liabilities. In India, the absence of such projections limits the assessment of fiscal risks and evolving pension commitments. 

Finally, better measurement should support — not precede —performance reforms. Attempts to link pay to performance often fail without a reliable and comprehensive measurement framework.   

India’s challenge is not simply whether to spend more or less on public sector compensation, but to understand what it is already spending. Without that clarity, fiscal policy risks underestimating long-term obligations. Improving how compensation is measured is, therefore, not a technical exercise. It is central to fiscal transparency, credible policymaking, and the effective functioning of the State.

The authors are with CSEP. The views are personal