This Selected Issues paper analyzes the medium-term macroeconomic outlook for India. The paper highlights that in the strong adjustment scenario, the joint effect of strong fiscal consolidation and ambitious structural reforms would bring the Indian economy onto a sustainable high growth path, reaching 7 percent around the turn of the century. Although there could be some initial dip in growth, owing to the contractionary effects of fiscal retrenchment, this should be temporary as a strong surge in investment, together with productivity improvements related to structural reforms, should drive an acceleration in growth performance.


This Selected Issues paper analyzes the medium-term macroeconomic outlook for India. The paper highlights that in the strong adjustment scenario, the joint effect of strong fiscal consolidation and ambitious structural reforms would bring the Indian economy onto a sustainable high growth path, reaching 7 percent around the turn of the century. Although there could be some initial dip in growth, owing to the contractionary effects of fiscal retrenchment, this should be temporary as a strong surge in investment, together with productivity improvements related to structural reforms, should drive an acceleration in growth performance.

IV. Controlling Central Government Expenditure 1

Reduction of the central government’s deficit since 1990/91 has occurred entirely on the expenditure side. This paper reviews the measures that have been taken to restrain spending over this period, and examines areas where there may be further scope for expenditure reduction. It focuses on four aspects of spending: wages and salaries, budgetary subsidies, grants and loans to states, and public expenditure management.2 It concludes that, while a sizeable reduction in non-interest current expenditure has been achieved since 1990/91, there nonetheless remains substantial further scope for improving the composition of expenditure through efforts to control unproductive spending. Restructuring and streamlining a number of expenditure programs would release resources for the Government’s priority areas, such as in infrastructure and the social sectors.

A. Trends in Expenditure

Between 1990/91 and 1995/96, the overall central government deficit was reduced by 2 ½ percentage points of GDP, entirely reflecting a lowering of the expenditure ratio. The brunt of deficit reduction fell on capital expenditure and net lending (down by 2.1 percent of GDP), notably in infrastructure and loans to states and public enterprises (Table IV.1). Non-interest current expenditure was reduced by 1.3 percent of GDP, but the impact was partly offset by rising interest payments.

Table IV.1.

Central Government Expenditure, 1990/91-1996/97

(In percent of GDP)

article image
Sources: Data provided by the Indian authorities; National Accounts Statistics (NAS); and staff estimates.

Including expenditure on goods and services.

Includes current and capital expenditure.

Includes current and capital expenditure in social and economic services.

Plan support in the form of equity and gross loans.

The reduction in non-interest current expenditure relative to GDP reflected contributions from all its main components:

  • Subsidies have been reduced by 0.7 percent of GDP, mainly due to the abolition of the export subsidy in 1991/92. Fertilizer and food subsidies have continued to be substantial.

  • Grants to states have been lowered by 0.7 percent of GDP, largely through cutbacks in plan grants.3

  • Wages and salaries (excluding the defense payroll) have been lowered by 0.4 percent of GDP, mainly through a 6 percent reduction in central government employment and—to a lesser degree—increases in real remuneration per employee that have lagged growth of per capita income (see Section B). However, the wage bill is now likely to be boosted substantially by the upcoming awards of the Fifth Pay Commission.

  • A decline of 0.3 percent of GDP in current expenditure on defense has similarly reflected slow growth of wages and salaries, which again may not be sustained given that the Fifth Pay Commission’s mandate includes an examination of pay conditions of defense personnel. Capital spending on defense has also been restrained; over the period, real spending on defense rose on average by 1.5 percent per annum, lowering total defense spending by 0.7 percent of GDP to 2.2 percent of GDP.

Cuts in capital expenditure and net lending occurred across the board. The impact on public investment was exacerbated by cutbacks (including in infrastructure) by states and public enterprises in response to the tightening of their budget constraints. With an attendant reduction in state spending in the social sectors, states’ development expenditure fell by 1 ¼ percent of GDP between 1990/91 and 1995/96. Public enterprises were able to cushion the impact in part through higher resource mobilization, but their plan spending has also declined—by ½ percentage point over the same period.

B. Wages and Salaries

Wages and salaries of the central government amounted to 1 ½ percent of GDP and almost 10 percent of total spending in 1995/96. The central government employs about four million people, about two thirds of whom are concentrated in two Ministries: the Ministries of Railways and Communication (Departments of Posts and Telecommunications) (Table IV.2). Including employment at the state level and central public enterprises, total public sector employment was about 13 million in 1993/94. Central government employment is relatively small because some of the most labor intensive government functions, such as education and health services, are primarily a state government responsibility.

Table IV.2.

Employment in the Central Government, 1990/91-1996/97 1/

(In thousands of persons)

article image
Sources: Union budget documents, various issues, data provided by the Indian authorities, and staff calculations.

Establishment strength on March 1. Data for 1996/97 are budget estimates.

Data include casual labor.

Civil service pay scales are effectively determined by the awards of the Pay Commission, a constitutionally mandated body set up to make recommendations about every five years on the central government’s pay scales, which are otherwise adjusted only for cost-of-living changes (the “dearness allowance”). These cost-of-living adjustments provide full compensation for employees in the lower pay scales, but only partial compensation for employees in the upper pay scales. As a result, pay scales tend to be compressed over time until restored with the Pay Commission awards. Interim awards are granted periodically, in anticipation of the release of the Commission’s recommendations.

Real remuneration per employee (which includes various allowances and fringe benefits) has increased on average by 1.8 percent per annum since 1990/91.4 However, the ratio of per employee compensation to per capita national income has declined by 4 percent—from 5.3 in 1990/91 to 5.1 in 1994/95—since real wages in government have lagged behind growth in all-India real incomes.

Since 1990/91, the Government has sought to reduce the size of the civil service by restricting recruitment below the rate of natural attrition.5 Specifically, in December 1991, the Prime Minister announced a target of reducing employment of all Ministries by 10 percent, although no specific time frame for the desired cut was given. Employment in the Railways has been reduced by 12 percent.6 Elsewhere, only a few-mainly small—departments were able to achieve the targeted cut by 1995/96. This mixed record reflects a reliance on mainly ad hoc measures, including:

  • Alignment of sanctioned posts with staffing needs of departments. A decision was taken in May 1993 that posts left vacant for one year would be abolished. However, in practice posts were often re-created during the annual budgetary exercise, undermining the impact of the measure.

  • Use of a pool of surplus employees to reduce the need for new recruitment. When departments need to fill vacancies or new posts, they are required to select employees from the surplus pool before looking elsewhere.7 However, with only few exceptions, unless a department or activity was eliminated, regular and systematic reviews were not conducted to identify surplus employees.8 As a result, the existing pool currently contains only 2,000-3,000 employees, most belonging in the “common category”, e.g., messengers, helpers, and drivers. Departments can generally frame job requirements such that they do not match the qualifications of employees in the pool.

  • Restriction on use of savings in wages and salaries for other departmental expenditures. A decision was taken in 1994 that savings in wages, salaries, and allowances resulting from vacancies could not be used for non-wage spending without approval of the Ministry of Finance. This was intended to control the number of sanctioned posts, given that each year’s estimates become the basis for incremental increases in the following year. No firm data exist on the effectiveness of this decision.

Looking ahead, the wage bill is likely to be boosted substantially by the implementation of the recommendations of the Fifth Pay Commission, which is expected to suggest a sizeable increase in pay scales. Although the recommendations have not been announced, the Government has already made a provision of Rs 50 billion (0.4 percent of GDP) in its 1996/97 budget to deal with the impact of the awards.9 An increase of this magnitude would return the wage bill to about the same ratio to GDP as in 1990/91. Beyond the present year, the Government has indicated its intention to freeze the real wage bill beginning in 1997/98, although detailed policies have not yet been specified (Ministry of Finance [1996]). Such a goal should be achievable without major retrenchment since there remains significant scope for achieving further savings through normal staff attrition in the civil service—by refraining from filling vacancies and enforcing vigorously the requirement to abolish posts left vacant for one year. Increasing private sector involvement, notably in telecommunications, will also create opportunities for a much needed restructuring of employment in the departmental enterprises. Reductions in employment would also generate savings in related expenditures (e.g., travel and telephone expenses, office supplies, as well as fringe benefits enjoyed by government employees),10 as well as, over the longer term, savings on pension liabilities.

A plan to reduce the size of government employment over the medium term should be part of a broader approach to civil service reform that would also address underlying problems such as lack of incentives and rewards for productive work. An overhaul of the organizational structure of the central government would help to streamline decision making, more closely match promotions to performance, and reduce duplication of responsibility.

C. Budgetary Subsidies

The main subsidies explicitly included in the central government’s budget are the fertilizer and food subsidies.11 There is considerable scope for reforming these schemes in order to achieve their principal objectives—promotion of agricultural production and income support—more effectively and at lower cost.

Fertilizer subsidy

The fertilizer subsidy was introduced in 1977 with the objectives of: (i) stimulating greater use of fertilizers to increase agricultural productivity; (ii) supporting agricultural incomes by providing cheap inputs; (iii) improving regional equity; (iv) achieving self-sufficiency in food grains; and (v) helping the domestic fertilizer industry establish self-sufficiency in fertilizer production based on the utilization of indigenous resources. Expenditure on the fertilizer subsidy declined modestly, from 0.8 percent of GDP in 1990/91 to 0.6 percent of GDP in 1993/94, reflecting a substantial reduction in the subsidy on non-nitrogenous fertilizers.12 However, the process was reversed in the 1996/97 budget, returning the fertilizer subsidy close to its 1990/91 level relative to GDP.

Nitrogenous fertilizers (urea) are currently subject to administered pricing and distribution. The subsidy on nitrogenous fertilizers is the difference between the retention price paid to the manufacturing units (public and private) and the consumer price of the fertilizer, less the distribution margin.13 A similar system of administrative controls on non-nitrogenous (phosphatic and potassic) fertilizers was eliminated 1992/93, but a subsidy was retained in the form of a fixed amount per ton paid to manufacturing units and importers for the sale of fertilizers on a concessional basis to farmers. This subsidy on the main phosphatic fertilizers was tripled from Rs 1,000 per ton to Rs 3,000 per ton in the 1996/97 budget.

The present system has a number of disadvantages, in addition to the direct budgetary cost:

  • The subsidy distorts economic incentives in the choice of input and output mix. This was particularly the case when the subsidy was concentrated on urea, which encouraged imbalanced nutrient use.

  • The subsidy is not well targeted: its benefits accrue mainly to better-off regions which have access to irrigation.

  • The Retention Price Scheme for urea rewards inefficient fertilizer producers and provides limited incentives for cost reduction. The urea industry in particular has a high cost structure, and retention prices are generally well above international prices.

A strategy to reduce the fertilizer subsidy would have to be part of a comprehensive plan to restructure pricing of both agricultural inputs and output and to improve rural infrastructure and credit delivery, which would enhance agricultural efficiency and address concerns about potential impact on income distribution. As suggested by the World Bank (1996), elements could include:

  • Phasing out of the retention price system for urea to eliminate the cost-plus-rate-of-return margin. Ultimately, prices of urea—like other fertilizers—would be free, and the industry delicensed. Nonviable public production units would be closed and profitable units privatized. In order to compensate for losses during the transition period, the Government would allow production units to become importers of fertilizers.

  • To cushion the burden on poor farmers, a ration card system could be introduced. Even if large farmers chose to participate in the scheme, the quantity of fertilizers used would be limited.

  • Supporting reforms of the agricultural sector would include freeing restrictions on internal and external commerce, which would improve the terms of trade for agriculture, and increased investment in irrigation and infrastructure in rural areas.

Food subsidy

The main objectives of the food subsidy system have been to: (i) protect farmer incomes through a minimum-support price; (ii) ensure national food security through the maintenance of buffer stocks by the Food Corporation of India (FCI); (iii) ensure adequate supply of food grains in all parts of the country through the Public Distribution System (PDS); and (iv) protect the vulnerable sections of the population from increases in prices of essential foodstuffs.

The FCI administers the purchase and distribution of foodstuffs at procurement prices fixed by the Government. The foodstuffs (mainly wheat, rice, and sugar) are primarily sold to the public through the PDS, at central issue prices fixed by the Government, using an extensive network of Fair Price Shops. All residents are in principle equally entitled to ration cards, although in practice access varies widely across the country, depending in large part on state-level policies. The difference between the economic cost (including operating costs and the carrying cost of buffer stocks maintained by the FCI) and the central issue price is reimbursed by the Government to the FCI.

Expenditure on the food subsidy has remained relatively constant at about 0.5 percent of GDP since 1990/91 (see Table IV.1). A temporary rise in 1993/94 reflected a combination of a good harvest and price adjustments that led to a 46 percent rise in procurement and a 14 percent drop in offtake during the year, adding to the costs of the FCI (Table IV.3). Increases in central issue prices outpaced increases in procurement prices from 1990/91 to 1993/94, as procurement prices increased modestly in real terms while central issue prices rose by 30 to 40 percent in real terms. Since February 1994, however, issue prices have not been adjusted.

Table IV.3.

Procurement and Central Issue Prices of Food Grains, 1990/91-1996/97

article image
Sources: Data provided by the Indian authorities; Food Corporation of India; and staff calculations.

The difficulty in reducing the food subsidy has reflected the conflicting objectives served by the minimum-support and central-issue price system, and the lack of effective control over the quantities of procurement and offtake. In particular:

  • Increases in support prices toward international levels have raised costs. Successive good harvests, the reluctance to export, and the resulting build-up of stocks have implied heavy support payments and carrying costs.14

  • As issue prices were increased, the differential from market prices was reduced, contributing to low offtake and escalating carrying costs.

  • The food distribution system is costly to administer. The FCI currently has about 65,000 regular and at least 35,000 contractual employees. A substantial part of the subsidy is accounted for by high wage costs and the carrying costs of buffer stocks maintained by the FCI. There are also substantial leakages in the PDS, including diversion of subsidized commodities to the market, which reduce the proportion of food reaching the poor.

As well as being costly, the PDS as presently structured has only limited effectiveness in protecting the poor. First, issue prices for wheat and rice are close to market levels. Second, the PDS is most extensive in areas that do not have direct access to locally produced food grains; it is thus regionally concentrated and tends to be focused in urban areas.15

The Government’s Common Minimum Program aims to strengthen the redistributive character of the system by providing essential foodstuffs to the poor at low prices, combined with better targeting to curtail costs. Although full details of the restructuring of the PDS and the associated costs have yet to be worked out, the Government has proposed the provision of food grains at half the market price to people below the poverty line. The categories of people to be excluded from the half-price scheme could include income-tax payers, those paying sales-tax registration fees, and government employees with salaries in excess of Rs 10,000 per month. However, experience in other countries has shown that targeting can be difficult to achieve, and the new scheme would need to be carefully structured to avoid a significant increase in the food subsidy bill.

An approach to reforming the food distribution system consistent with the Government’s objectives could combine: (i) determined efforts to reduce the costs of administering the PDS and the escalating costs of carrying stocks; and (ii) greater use of self selection as a basis for improved targeting.

  • Operational costs of the PDS could be substantially reduced through decentralization of the FCI and increasing reliance on private distribution channels. Leakages at the PDS could be cut by improving monitoring, supervision, and by introducing accountability, as well as by selling only lower quality grains which are less likely to be diverted to the market. Moreover, the need to maintain a high internal buffer stock could be reexamined, in light of increasing integration of international markets and diminishing food security concerns.

  • Targeting of the PDS could be improved by reducing incentives for high income persons to use the PDS, for example, by distributing only low quality grains consumed mainly by the poor and discontinuing the sale of sugar and finer grades of wheat and rice from PDS shops. Self selection would be enhanced by locating PDS shops in poor areas only, and by expanding food-for-work programs.

D. Grants and Loans to States

Under the existing federal structure, central government grants and loans to states consist of statutory (mandatory) and non-statutory (discretionary) transfers.16 The main statutory transfers are: (i) non-plan grants determined on the basis of the recommendations of the Finance Commission, which in the past have been set at levels designed to cover projected revenue deficits (i.e., deficits in current operations) of states; and (ii) non-plan loans against small savings collections, equal to 75 percent of net collections in the state. The most important non-statutory transfers are grants and loans as central support to state plans and to central or centrally sponsored schemes.17 Grants for central support to state plans are unconditional, while grants for central or centrally sponsored schemes are given for the implementation of specified development or welfare programs. The distribution of central support to state plans among the different states is governed by a specific formula (the modified Gadgil formula), although the total amount of central support to state plans is at the discretion of the central government, under the guidance of the Planning Commission.18

Total transfers to states have declined by 1 ½ percent of GDP since 1990/91, reflecting a reduction in both statutory and non-statutory transfers (Table IV.4). The reduced availability of central resources to the states, and in particular the cutbacks in plan grants and loans, has prompted a sharp reduction in capital spending at the state level.19 States have also sought to boost their access to resources by providing additional incentives for collections of small savers deposits, which explains the rise in non-plan loans to states in 1994/95 and 1995/96.

Table IV.4.

Transfers to States, 1990/91-1996/97 1/

(In percent of GDP)

article image
Sources: Data provided by the Indian authorities; and staff calculations.

Exclude states’ share of centrally collected taxes.

Small savings schemes have created a number of problems. First, they have increased the indebtedness of states to the central government, and some states have had increasing difficulties in servicing these loans. Second, the sizeable implicit subsidy of the scheme has imposed a significant fiscal burden on the central government.20 In an effort to curtail the growth in these deposits, the central government directed in 1994/95 that financial institutions not be allowed to make such deposits. Further options to reduce intermediation costs would include charging an interest rate on loans against small savings collections more in line with costs, and reducing the maturity of these loans.

A number of Government initiatives are expected to affect the composition and size of central transfers to states. In particular:

  • The Government proposes in its Common Minimum Program to delegate greater spending responsibility to states. These proposals would entail granting states greater autonomy in drawing up their plans and determining their development priorities, transferring most centrally sponsored schemes to the control of state governments, and re-examining the devolution of financial powers from the central government to state governments.

  • Implementation of the Tenth Finance Commission’s recommendations would encourage fiscal discipline at the state level, and provide a stable source of revenue. These recommendations include: (i) the phasing out by 1999/2000 of non-plan grants that serve to cover projected deficits in current operations of states; (ii) provision of debt relief to states linked to good fiscal performance, and additional write-off of debt equivalent to the debt retired by states using proceeds from divestment of equity in state enterprises; and (iii) the sharing of central taxes with states from a single divisible revenue pool at a uniform rate.21

E. Public Expenditure Management

In principle, India’s expenditure control and management system includes performance budgeting, complete separation of accounting and auditing, full internal controls in the spending agencies, zero-base budgeting, internal audit and management information systems, as well as year-end evaluation of plan programs and projects. However, in many respects, these systems have not proved effective. Some of the procedures have not become fully operational (e.g., performance budgeting), while others are not used in practice (e.g., zero-base budgeting). As a result, measures adopted since 1990/91 to control spending have frequently relied upon ad hoc, across the board cutbacks rather than the systematic reassessment of program needs and priorities.22

The new Government has taken a number of steps to control spending and improve public expenditure management across a number of fronts, as described in a circular issued in June 1996 (Ministry of Finance [1996]). The measures included:

  • Review of programs by ministries and identification of redundant schemes by September 1, 1996. A special unit was created in the Department of Expenditure to monitor progress on the pruning of schemes, to be reflected in the 1997/98 budget.

  • Monthly review of receipts and expenditures and preparation of monthly cash profiles for each program to improve cash management.

  • An end to the practice of carrying forward of liabilities to the next fiscal year to ensure that budgetary appropriations in the following year have proper authorization.

  • All profit-making public enterprises have been required to distribute a minimum dividend on equity of 20 percent, or a minimum dividend of 20 percent of post-tax profits, whichever is higher, to provide a reasonable return on the capital employed. All profit making enterprises are also required to issue bonus shares to the Government.

  • With a view to reducing government employment, there will not be any real increase in budgetary allocations for pay and allowances beginning in 1997/98. Each ministry and department are required to provide at the time of formulation of their budget estimates a plan for reduction in their staff.

  • In addition, the formation of an Expenditure Commission was announced in the 1996/97 budget with a mandate to review programs and expenditure procedures in the central government. Additional public expenditure management reforms that the Expenditure Commission could consider include:

  • Eliminating the distinction between “plan” and “non-plan” expenditures, which encourages a segmented approach to expenditure management and has resulted in little medium-term planning of non-plan spending.

  • Introduction of rolling medium-term expenditure plans, which would reveal the full expenditure implications of existing programs.

  • Creation of expenditure profiles, which would identify the sources of program cost escalation.

  • Development of performance indicators and targets, which would permit an assessment of a program’s economic and social returns.

  • Introduction of firm cost limits for spending agencies, to facilitate more effective cash management and make short-term borrowing needs more predictable.

  • Introduction of an integrated general ledger system and computerized budget management system.


  • Ministry of Finance (1996), Guidelines on Expenditure Management—Fiscal Prudence and Austerity, Office Memorandum of the Department of Expenditure (India).

    • Search Google Scholar
    • Export Citation
  • World Bank (1995), India—Country Economic Memorandum, Economic Developments in India: Achievements and Challenges.

  • World Bank (1996), India—Country Economic Memorandum, Five Years of Stabilization and Reform: The Challenges Ahead.


Prepared by Dimitri Tzanninis.


The paper makes only selected reference to spending at the state level. A more detailed discussion of state-level expenditure issues is contained in World Bank (1995), and in Chapter I of SM/92/205 (1½3/92). Furthermore, Chapter III of SM/94/48 (2/25/94) illustrated the issues involved in state finances with a case study of Uttar Pradesh.


Grants to states constitute the single largest category of non-interest current expenditure of the central government. In 1995/96, about 70 percent of these grants were plan grants, which are block grants given as central support to state development plans.


This increase masks a gradual erosion in real wages in the upper pay scales of the civil service through the partial indexation for cost-of-living changes.


About 3 percent of civil servants leave the payroll each year.


It is not clear how much of the reduction in employment in the Railways represents reductions in casual labor.


Surplus employees receive foil compensation, and are required to report to work, but have no duties to perform. Although surplus employees cannot object to a transfer, it has been very difficult in practice to redeploy employees in areas outside the area of their recruitment. Moreover, there is no retraining to improve mobility of surplus or regular employees.


A reliable estimate of surplus labor in the central government would require a manpower audit, which does not exist in India. The manpower audit is needed to review functions of departments, and identify surplus labor and redundant or overlapping departments.


Previous Pay Commission recommendations have typically included some retroactive component.


Such benefits include, among others, subsidized housing, education, and health services.


The previously large export subsidy (0.5 percent of GDP), was eliminated in 1991/92 in the context of trade liberalization and exchange market reform. There are a range of other subsidies, many substantial, in the Indian economy, including on rural electricity and irrigation, kerosene and liquefied propane gas, passenger rail traffic, and small loans. Many of these are covered by cross-subsidization within a sector or incur a burden on state, rather than central government, finances.


The size of the subsidy understates the true budgetary cost, since public production units as a whole suffer losses, which are covered separately through budgetary transfers. Three of the nine public production units are chronic loss-makers, with losses totaling Rs 8.9 billion (0.1 percent of GDP) in 1995/96.


A pre-tax return on net worth corresponding to post-tax return of 12 percent is given to manufacturers under the Retention Price Scheme. Retention prices vary by manufacturer, depending on their unit costs.


Grain exports, however, increased in 1995/96 in response to good supply conditions and large stocks.


The Revamped Public Distribution System (RPDS), introduced in 1992, is an exception to this general feature of the system, providing foodstuffs to tribal and inaccessible areas, where the incidence of poverty is higher, at prices substantially lower than in PDS shops.


In addition to grants and loans, statutory transfers include a share of some centrally collected taxes distributed according to the recommendations of the Finance Commission. See SM/92/205 (1½3/92).


Centrally sponsored schemes are designed and monitored by the central government, implemented by state governments, and financed mainly by the central government.


The modified Gadgil formula allocates central assistance to states on the basis of fixed weights assigned to population, state per capita income, fiscal management, and state-specific problems.


States have limited authority to raise financing on their own account. The Constitution forbids states from accessing foreign financing, and from borrowing on domestic markets without the central government’s consent if there are any outstanding liabilities to the central government.


The interest rate charged on these loans is 14.5 percent per annum. However, the effective cost to the central government from administering the small savings scheme is about 18 percent per annum on these funds. Loans against small savings collections have a maturity of 25 years and a five-year grace period, although the deposits themselves are only up to five years in maturity.


The Tenth Finance Commission has recommended that the states’ share of revenues collected by the central government be set at a uniform rate of 29 percent and frozen for 15 years. Customs duties, corporate income tax, and services tax-which at present are not shared with states-would be included in the divisible pool, along with the currently shared excise duties and personal income tax.


These measures included the requirement that posts left vacant for one year would be abolished, the decision to cut employment and departmental expenses by 10 percent, and strict enforcement since 1994 of an older rule that departments and agencies could not transfer funds from plan to non-plan expenditure.