By 2004-05, as compared to 1950-51, the compensation structure of central government employees had gravitated substantially towards deferred payments. Though the average wage compensation of a central government worker is higher than the per worker gross domestic product, the latter has grown at a faster rate in the past few years. The number of central government workers has been declining and the aggregate expenditure towards employee compensation as a proportion of net non-debt revenue receipts has also fallen. Hence, at the aggregate level, there is no grave concern regarding the fiscal implications of wage and retirement benefits

MUKESH ANAND, SASWATA CHAUDHURY

T

A cursory look at the terms of reference (ToR) of the SPC reveals some discomforting features (see Annexure A). Some of the terms, especially B, obfuscate the role of the commission while certain others (like C and D) burden the commission with avoidable jargon and yet others (like E and F) bind the commission to a strait-jacketed approach favouring generalisation, as opposed to incentivising specialisation. A commission set-up every decade or so, may not be the best conduit “to transform the central government organisations into modern, professional and citizen-friendly entities dedicated to the service of the people” (see Annexure A, ToR B). This perhaps, should be a continuing endeavour/activity. Further, the terms C, D and F have redundant or superfluous clauses. For example, “…comprehensive package…”, “…totality of benefits…”, in C and D, or, “promoting efficiency”, “productivity linked incentive scheme” in D and F, etc, that ostensibly describe similar intents. A simple enunciation of objectives and ToR is desirable in advancing monitorability of outcomes and identifying the appropriate deliverables without obscuring the purpose of such commissions.

At a macroeconomic level, such commissions provide us with an opportunity to debate upon the desirable size and composition of government establishments and the level of wage price in the cost of service rendered. In economic terms this is equivalent to a technological choice for labour (in the production of government/public services) and the corresponding factor income (wages). A distinctive characteristic, however, is that unlike the usual (market) processes where demand and supply interact to produce the price signal (level), it is the converse mechanism that occupies centerstage. This is analogous to the erstwhile administered price mechanism in certain primary sectors. The pay commission is entrusted with an important exercise to provide this key ingredient that profoundly influences the economy of the state in its monopsonistic position. The extant commission is sixth in the series of central pay commissions. Table 1 lists the dates of appointment of successive commissions.

At the microeconomic level, keeping wages depressed cannot be presumed to translate into lower costs (for the populace) of government (public) service or be a comprehensive indicator of efficiency and productivity. Moreover, there ought to be sufficient dis-incentives to prevent any dilution in the hierarchical relationships for smooth functioning of the government (both, across the public and private sectors as well as within the public sector). But governments in developing economies cannot be perceived to foster wide inequalities. The ToR could be thus geared to ensure not only sufficient financial capacity for a decent living, but also sufficient opportunity for decent work [ILO 2001, especially Chapter 2].

Section I, tracks the central government (CG) expenditure on wages and salaries, as well as on pensions and retirement benefits. This is discussed in relation to net non-debt revenue receipts (NNDRR) and GDP at the aggregate level. In Section II, we discuss the changes in CG employment and the corresponding per employee emoluments. These are compared against nominal

Economic and Political Weekly August 4, 2007

Figure 1: Structure of Employees’ Cost to the Establishment ‘Economic and Functional Classification of the Budget’ (EFCB)

Aggregate Wage and Pension Expenses

Over the period 1950-51 to 2004-05, wage expenditure for

Compensation Structure(per cent) 100 95 90 85 80 75 70 65 60

general government administration (GGA) has grown by 12.43

per cent per annum, while that for DCUs has grown by 11.06

per cent per annum. The combined expenditure on wages for GGA and DCUs has grown at the rate of 11.92 per cent per annum. In addition, the expenditure on retirement benefits, for both GGA and DCUs together, has grown at the rate of 17.01 per cent per annum over the same period. Thus, the total cost of government

1950-51

1956-57

1962-63

1968-69

1974-75

1980-81

1986-87

1992-93

1998-99

2004-05

2004-05

Year

Pension

Wages

employment (wages and salaries, and pensions) grew at a trend

rate of 12.55 per cent per annum between 1950-51 and 2004-05. Table 2 summarises the trend rate of growth in wage, pension and total expenditure over the periods between the successive

Source: Same as in Table 2.

Figure 2: Employee Compensation and Net Non-Debt Revenue Receipts

(Per cent) pay commissions.8 The differential rates of growth of the two components, namely

EC/NNDRRNNDRR/GDP 80 60 40 20 11 9 7 5 3

wages and pensions, have resulted in significant restructuring

of the employee cost of government establishment. Figure 1

tracks the changes in this structure. Wages constituted 92.05 per

cent of the total compensation in 1950-51, but this proportion

declined to 69.43 per cent in 2004-05. Conversely, retirement

benefits that constituted less than 8 per cent of total employees’ compensation (EC) in 1950-51 have now risen to 30.57 per cent

in 2004-05.

Such a structure is quite likely to have a perverse influence on incentives and motivation for labour and effort, if a continuously growing (and in certain cases like defence, a larger)

1950-51

1956-57

1962-63

1968-69

1974-75

1980-81

1986-87

1992-93

1998-99

Year

EC/NNDRR (LS)

NNDRR/GDP (RS)

component accrues as deferred compensation.

Notes: EC: Employees Compensation; NNDRR: Net Non-Debt Revenue Receipts;

GDP: Gross Domestic Product; LS: Left-hand Scale; RS: Right-hand Scale. Source: Authors’ own computations; Basic Data: EC, same as in Table 2;

NNDRR [GoI 2006b]; GDP [GoI 2006c].

and real changes in per worker GDP (PWGDP). Section III summarises the findings.

The central government establishment (CGE) encompasses all personnel engaged with its ministries, (independent) departments, departmental commercial undertakings (DCU) and statutory commercial corporations. While no clear line can be drawn on what constitutes this behemoth (almost all the commissions have made some remarks on defining central government), the working definition alludes to all employees who draw their wages and salaries and, superannuation and retirement benefits4 from the Consolidated Fund of India (Report of FPC, Vol 1, p 7). As per the central government budget papers, there were 3.23 million civilian workers in 2005-06.5 This transforms to 2.9 central government civilian workers per 1000 persons in the population. Guess estimates place the number of defence workers (in military services) at roughly 30 per cent of the number of civilian workers.6 Available estimates on the number of pensioners, of central government services, are less reliable, and some related issues are discussed in the next section.7 In this section we analyse the growth and structure of aggregate employee compensation, basic data for which is compiled from various issues of the

II Per Worker GDP and Average Compensation of CG Employees

In the previous section we discussed the structure of compensation at an aggregate level. Deciphering the factors contributing to such a structure is the subject matter of subsequent research, while presently we take a closer look at average compensation.14 In the first sub-section we estimate the average GDP per worker followed by a sub-section each on average compensation accruing to current and retired employees in CG services.

Number of Workers and GDP Per Worker

Data relating to GDP has been set out in the previous section, and information on the number of workers has been collated from two sources15 namely, the census reports (Registrar General and Census Commissioner, India)16 and the National Sample Survey Organisation (NSSO).17 Unfortunately, the years for which data could be collated from the two sources are different. Table 5 summarises the proportion of working persons in India from these sources.18 The estimates of per worker GDP, at constant 19992000 prices, are also presented.

Columns 1-5 (in Table 5), correspond to basic data from the NSSO and present estimates of PWGDP for two differing considerations, that the working population constitutes (a) those with principal and subsidiary worker status (columns 2 and 4) and, (b) those with principal worker status only (columns 3 and 5). Columns 6-8 correspond to basic data from the censuses. While estimates based on NSSO data in columns 4 and 5 suggest

Table 5: Proportion of Working Population and Per Worker GDP Figure 3: Central Government Workers and Per Worker Annual Compensation (at constant 1999-2000 prices)

(at 1999-2000 Rs)

	NSSO			Census
Year	Working Population (Per Cent)	PWGDP (Rs)	Year	Working Population	PWGDP
	PS+SS	PS	PS+SS	PS		(Per Cent)	(Rs)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
1977-78	42.3	37.1	23618	26928	1961-62	42.96	18702
1983-84	42.0	37.4	26321	29559	1971-72	32.93	28397
1987-88	41.2	36.5	29599	33411	1981-82	36.70	28715
1993-94	42.0	37.5	35355	39597	1991-92	37.50	37935
1999-2000	39.7	36.5	49121	53427	2001-02	39.10	53208
2003-04	42.1	38.1	53178	58762
Notes:	PWGDP: Per Worker GDP at constant 1999-2000 prices; PS: Principal Status; SS: Subsidiary Status.
Source: Authors’ own computations; Basic Data, NSSO, Census of India.
Economic and Political Weekly	August 4, 2007		3227

Panel A: GGA Panel B: CGE

Compensation('000 Rs) 120 95 70 45 20 Workers (lakhs) 35 30 25 20

Compensation('000 Rs) 100 75 50 Workers (lakhs) 42 38 34 30

1961-621969-701977-781985-861993-942001-02

1987-88

1991-92

1995-96

1999-2000

2003-04

Year Year Compensation Workers

Workers Compensation

Notes: GGA, General Government Administration; CGE, Central Government Establishment; One lakh equals 100 thousand.

Source: Authors’ own computations; Basic Data, GoI 2006a, b, c, 2007b, 2005b; Panel A utilises data from DGE&T and the EFCB; Panel B utilises data from Annexure VII in the Budget.

a near doubling of PWGDP in two decades between 1983-84 and 2003-04, this is not fully corroborated by figures from the Census for the years 1981-82 and 2001-02 (column 8).19

Choosing the broader definition for a worker and utilising the estimates presented in column 4, PWGDP grew at a compounded annual average rate of 3.73 per cent per annum between 198788 and 2003-04.20 However the rate was not uniform over the 16-year interval. Data limitations prevent us from making a precise comparison between the pre- and post-FPC periods, but Table 6 depicts the compounded annual average growth rate (CAGR) for two combinations of intervals.

In contrast, the CAGR of PWGDP for the six-year period between 1993-94 and 1999-2000 was significantly higher at 5.63 per cent per annum. In the next sub-section we take a closer look at the average compensation level of a worker with CG.

Central Government Workers and Per Worker Compensation

In comparison to the intermittent information available on allworkers in India, data for the number of workers in the public sector, specifically CG, is available for the last several years (see footnote 15). However, data with consistent coverage across descriptions and dimensions are not easily available in the public domain. This issue is lamented in each of the earlier reports of the central pay commissions. For example, DGE&T gives the number of workers in the public sector classified under central government establishment and quasi central government establishment (Q-CGE).21As per the description of data from DGE&T, workers in military services are included under CGE and, those employed with DCU and statutory corporations are grouped under Q-CGE.22

Next, Annexure VII of the Expenditure Budget (Volume I) of the union government gives some details on the number of workers and cost of government establishment by each ministry, but pertains only to civilian employees and does not segregate the data for DCUs.23 Table 7 summarises the sources and their scope of coverage relevant for this study.

Figure 3 gives the estimated number of workers and per worker compensation in CG at constant 1999-2000 prices.24 As per the budgets (Figure 3, Panel B), the strength of CGE declined at the rate of 1.47 per cent per annum between 1988-89 and 2003-04.25 The decline is significantly steeper in the post-FPC period between 1996-97 and 2003-04, at 2.46 per cent per annum. Over the period 1988-89 to 2003-04, per worker compensation grew at the rate of 3.61 per cent per annum, while in the post-FPC period, between 1996-97 and 2003-04 (2005-06) it was 2.68 (2.46) per cent per annum.26This trend rate of annual growth realised in the post-FPC

Table 6: Growth of Per Worker GDP

(Per cent per annum)

Period Rate Period Rate

1987-88 to 1993-94 (6) 3.01 1987-88 to 1999-2000 (12) 4.31 1993-94 to 2003-04 (10) 4.17 1999-2000 to 2003-04 (4) 2.00

Notes: Figures in parentheses give the length of the period. Source: Same as for Table 5.

Table 7: Sources and Coverage of Data on Employment andEmployees Compensation, in the Public Sector

Scope

Civil Military DCUs Statutory

Source

Corporations

Directorate general of
employment and	CGE	√	√
training (workers)	Q-CGE			√	√
Expenditure Budget
(vol I) – Annexure VII
(workers and wages
and salaries)		√		√
Economic and
Functional Classification
of the Budget (wages
and salaries and,	GGA	√	√
pension)	DCUs			√

Notes: CGE: Central Government Establishment; Q-CGE: Quasi-Central Government Establishment; GGA: General Government Administration; DCU: Departmental Commercial Undertaking.

Source: Authors’ own comprehension; GoI (2007a, 2006a, 2005a, b).

Table 8: Groupwise Range of Annual Increments

(Per cent)

Group	Range
A B C D	2.24 – 3.05 2.35 – 2.99 2.00 – 2.31 1.42 – 1.54

Notes: Annual increments expressed as a percentage of the mid-value for each of the scales in each group.28

Source: Authors’ own computations; Basic Data, Report of the FPC.

period lies within the range of annual increments (that serve as a counter check) implicit in the award of FPC, summarised in Table 8.27

Putting together the values from Figure 3 and Table 5, one observes that the average compensation of CG workers is significantly higher than PWGDP. However, growth in PWGDP

(3.73 per cent per annum between 1987-88 and 2003-04) is marginally higher than growth in average compensation (at 3.61 per cent per annum between 1988-89 and 2003-04).29

The awards following the last two pay commissions were implemented from a retrospective date, leading to lumpy arrear payments. Thus, the estimate of differential between the trend rates of growth (of PWGDP and average compensation to central government workers), as also the trend rates of growth (of average compensation to central government workers) are sensitive to the choice of interval or a sub-interval. This is apparent when we, alternatively, utilise the data from the DGE&T and, the EFCB (Figure 3, Panel A). The real increase in average compensation to a worker, engaged in GGA, clocked a trend rate of 4.31 per cent per annum between 1988-89 and 2002-03. While the increase in the post-FPC period between 1996-97 and 2002-03 was 4.28 per cent per annum, the corresponding increase between 1988-89 and 1996-97 was a mere 1.23 per cent per annum.30

The differences in the results, from the two sources, may also be due to differences in the scope and coverage of constituents. As Table 7 reveals, Annexure VII (of the budget) data includes expenditure on workers of DCUs but excludes that on the military, and these perhaps differ significantly in certain characteristics. For example, presently, the number of workers with the DCUs exceeds that with the military. Next, the structure of workers (the distribution over groups A, B, C and D) in military is perhaps significantly different from that for the DCUs (in turn influencing the level of average compensation). In recent years perhaps the rate of new recruits into DCUs has slowed while the rate of retirement has accelerated. In contrast the strength of workers in the military has remained largely unchanged. Moreover, there are differences with regard to length of normal service and

Table 9: Number of Pensioners and Pension Expenditure: GGA

Year Pensioners Total Average Replace-System (Lakhs) Pension Pension ment Ratio Depen-Expenditure (Rs) (Per Cent) dency (Rs Crore) Ratio (Per Cent)

(1) (2) (3) (4) (5) (6)

1990-91 18.53 2138.23 25223 38 54 1994-95 20.58 3653.74 26372 40 61 1999-2000 25.23 14285.92 56623 56 77 Compounded annual average growth rate (CAGR, per cent) 1990-91 to

1994-95 (4) 2.66 14.33 1.12 1.29 3.09 1994-95 to

1999-2000 (5) 4.16 31.35 16.51 6.96 4.77 1990-91 to

1999-2000 3.49 23.50 9.40 4.40 4.02

Notes: Pensioners include family pensioners and other survivor pensioners. Average pension is pension amount at constant 1999-2000 prices. Replacement ratio here is the average payment to a pensioner, from GGA, as a ratio of average payment to a current worker (unlike the usual definition pertaining to the ratio of monthly compensation drawn by a given individual upon retirement in relation to his/her own monthly wages while in service). System dependency ratio is the number of pensioners from GGA as a ratio of the number of current workers in GGA (unlike age-dependency ratio or worker-dependency ratio).

Source: Report of the FPC, Report of the High Level Expert Group, Union Budgets, Finance Accounts.32

average age of new entrants in military and civilian services [see for example, Unnithan 2007].

Between 1950-51 and 2004-05, nominal GDP at market prices grew at a trend rate of 12.13 per cent per annum. This, coupled with buoyant revenue resources of the union government has yielded a 13.76 per cent per annum growth in NNDRR. The cost of CG employees’ compensation (workers and retirees), expressed as a proportion of NNDRR, has halved from around 60 per cent (in 1950-51) to less than 30 per cent (in 2004-05). Thus, at the macroeconomic (aggregate) level, there is no grave concern with regard to the fiscal implication of wage and retirement benefits of CG employees.

Analysis at the microeconomic level is constrained by data availability and incongruity in coverage of the various sources of data. One may however generalise that over the last 15 years or so, per worker GDP has grown marginally faster than average compensation of central government workers. Average compensation of CG workers though is significantly higher than PWGDP. Whether the average is sufficient to provide for a decent standard of living remains a moot concern. The pay commission needs to strike a delicate balance so that the legitimate concerns, including the need to optimise cost of public services while preserving incentives for labour and effort, are adequately addressed.

Presently, there are almost 3.9 CG (civil plus defence) workers per 1,000 persons in the population, and a smaller (but rapidly rising) number of retirees or ex-workers (from CG services). The number of workers in civilian services has been declining at a trend rate of 1.47 per cent per annum, over the period 1988-89 and 2003-04. The downsizing of CGE accelerated in the post-FPC era. In contrast, the number of retirees from GGA (that includes defence but excludes DCUs) has grown at a trend rate of 3.49 per cent per annum during the 1990s. The trend accelerated

Economic and Political Weekly August 4, 2007 in the post-FPC period. Expenditure towards wages and salaries of the current employees and retirement benefits of ex-employees from CG services impinges on the Consolidated Fund of the union government. Compensation of current employees in 1950-51 constituted more than 92 per cent of employees’ costs to the CGE. By 2004-05, current compensation had declined to less than 70 per cent with remaining being claimed as deferred compensation or retirement benefits of ex-workers.

The rapid rise in retirement benefits is perhaps a cause of concern. In 1999-2000, average retirement benefits to exemployees from GGA had reached 56 per cent of average wage compensation of current employees. It is likely that this has risen higher in the intervening years up to the present. The absence of an appropriate mechanism to adjust the service and retirement rules that incorporate evolving demographic and social developments appears to be the major drawback. An obvious lacuna pertains to inadequate synchronisation of the normal age of superannuation with improvements in expectation of life. This is likely to be the most important factor contributing to the extant structure of current and deferred compensation (over wages and pensions). This has worsened not only the passivity ratio34 but also the system dependency ratio [Anand 2007]. The latter ratio in central government services is hovering around 80 per cent, that is, almost eight retirees (ex-workers) for every 10 workers. The key factors influencing the structure of compensation and likely to be of concern for the pay commission constitute the subject matter for subsequent research.

Annexure A: Terms of Reference for the Sixth Central Pay Commission

(A) To examine the principles. The date of effect thereof that should govern the structure of pay, allowances and other facilities/benefits whether in cash or in kind to the following categories of employees:

(1) Central government employees – industrial and nonindustrial.

(2) Personnel belonging to the all India services.

(3) Personnel belonging to the armed forces.

(4) Personnel to the union territories.

(5) Officers and employees of the Indian audit and accounts department; and

(6) Members of the regulatory bodies (excluding the RBI) set up under Acts of Parliament.*

(B) To transform the central government organisations into modern, professional and citizen-friendly entities that are dedicated to the service of the people.

(C) To work out a comprehensive pay package for the categories of central government employees mentioned at (A) above that is suitably linked to promoting efficiency, productivity and economy through rationalisation of structures, organisations, systems and processes within the government, with a view leveraging economy, accountability, responsibility, transparency, assimilation of technology and discipline.

(D) To harmonise the functioning of the central government organisations with the demands of the emerging global economic scenario. This would also take into account, among other relevant factors, the totality of benefits available to the employees, need of rationalisation and simplification, thereof, the prevailing pay structure and retirement benefits available under the central public sector undertakings, the

economic conditions in the country, the need to observe fiscal prudence in the management of the economy, the resources of the central government and the demands thereon on account of economic and social development, defence, national security and the global economic scenario, and the impact upon the finances of the states if the recommendations are adopted by the states.

(E) To examine the principles which should govern the structure of pension, death-cum-retirement gratuity, family pension and other terminal or recurring benefits having financial implications to the present and former central government employees appointed before January 1, 2004.

(F) To make recommendations with respect to the general principles, financial parameters and conditions which should govern payment of bonus and the desirability and feasibility of introducing productivity linked incentive scheme in place of the existing ad hoc bonus scheme in various departments and to recommend specific formulae for determining the productivity index and other related parameters.

(G) To examine desirability and the need to sanction any interim relief till the time the recommendations of the commission are made and accepted by the government.

*A.6 substituted by Ministry of Finance Resolution No 5/2/ 20006-E.III (A), December 7, 2006. The commission will devise its own procedure and may appoint such Advisers, institutional consultants and experts, as it may consider necessary for any particular purpose. It may call for such information and take such evidence, as it may consider necessary. Ministries and departments of the government of India will furnish such information and documents and other assistance as may be required by the commission. The government of India trusts that state governments, service association and others concerned will extend to the commission their fullest cooperation and assistance.

The commission will have its headquarters in Delhi.

The commission will make its recommendations within 18 months of the date of its constitution. It may consider, if necessary, sending reports on any of the matters as and when the recommendations are finalised.

m

Email: saswata.chaudhury@gmail.com

6 As per the data collected under the employment market information programme of the Directorate of Employment and Training [GoI 2005a] in 2002-03, there were 6.31 million central government workers. Of these,

7 As per the report of the FPC [GoI 1997: 1826], there were 3.23 million pensioners (from central government services) in 1994-95 (including million 1.67 million defence pensioners). Anand and Ahuja (2004) utilise

1.90 million and 1.94 million for the number of civilian and defence pensioners respectively, for April 2001. 8 As far as practicable, analysis in rest of the paper follows identical period classification. 9 Net of statutory transfers, to states, as per the award of the successive Finance Commissions and, excludes capital receipts.

10 Until 1979-80, occurrence of revenue deficit in the centre’s budget was a rarity but has been a permanent feature ever since [Anand, Bagchi and Sen 2004]. One can therefore easily deduce that, over the period 1950-51 to 2004-05, employees’ compensation has seen a steeper decline as a proportion of centre’s revenue expenditure.

11 Thus the share of taxes in the net receipts of the central government has declined by more than 5 percentage points since the beginning of 1950s.

12 The detailed steps in constructing the 1999-2000 series of GDP at current prices are as follows: The GDP values in 1993-94 series are available up to 2003-04; The GDP values in 1999-2000 series are available from 1999-2000 onwards; the common years in the two series (namely, 1993-94 and 1999-2000) are between 1999-2000 and 2003-04; We estimate the ratio of GDP values in 1993-94 series to GDP values in 1999-2000 series for each of the common years; The average of this ratio for the five common years is 0.996; We use this as an adjustment factor to convert the past GDP values in 1993-94 series, into 1999-2000 series; Thus for example, GDP in 1970-71 in 1999-2000 series is GDP of 1970-71 in 1993-94 series divided by 0.996. Next, at constant 1999-2000 prices, value of GDP in 1999-2000 series is generated by multiplying the value of GDP in 1993-94 series at constant 1993-94 prices with 1.54 (where 1.54 is the ratio of GDP at constant 1999-2000 prices to GDP at constant 1993-94 prices for the year 1999-2000).

13 Between 1950-51 and 2004-05, nominal GDP at factor cost grew at a trend rate of 12.01 per cent per annum while real (at constant 1999-2000 prices) GDP at factor cost grew at 4.38 per cent per annum.

14 However, the average may not be the best measure of central tendency.

15 Another source namely, The Labour Bureau of the Directorate General of Employment and Training (DGE&T), Ministry of Labour (MoL), Government of India [GoI 2005a, b] also provides regular data (although with a lag of four to five years) on organised or formal sector employment only. The data given is collected under the Employment Market Information Programme of the DGE&T. However, the coverage, even of the organised sector, is less than complete. It includes all establishments in the public sector irrespective of their size (excluding employment in Indian missions abroad and part-time staff or persons employed through contractors) and non-agricultural establishments employing 25 or more persons in the private sector statutorily under the provisions of the employment exchanges (compulsory notification of vacancies) Act, 1959. Commencing from March 1966, the smaller establishments in the private sector employing 10-24 workers are covered on a voluntary basis.

16 The census surveys are conducted every 10 years.

17 We have utilised data from the quinquennial rounds of NSSO.

18 As per the census a worker is defined as one who has participated in any economically productive activity at any time during the reference period. The NSSO classifies workers into two groups by (a) principal status and (b) subsidiary status. Activity status on which a person spent relatively longer time (i e, major time criterion), during the 365 days preceding the date of survey is considered as the usual principal activity status of the person. A person, whose usual principal status was determined on the basis of the major time criterion, could have pursued some economic activity for a shorter time throughout the reference period of 365 days preceding the date of survey or for a minor period, which is not less than 30 days, during the reference period. This is classified as the subsidiary economic activity status of that person.

19 Combining the estimates using census and NSSO data, it may be summarised that PWGDP grew rapidly in the 1960s (4.26 per cent per annum from 1961-62 to 1971-72), but the rate of growth perhaps slowed down considerably towards the end of the 1960s and up to the beginning of the 1980s (0.11 per cent per annum between 197172 and 1981-82). After a brief period of acceleration in the second and third quarter of the 1980s (averaging above 3 per cent per annum), the rate of growth of PWGDP decelerated again, but resurged after the first quarter of the 1990s. This was however short-lived and the rate of growth of PWGDP started to decelerate towards the end of the 1990s and in the beginning of the 21st century (2 per cent per annum between 1999-2000 and 2003-04). The situation has perhaps reversed dramatically in the most recent years, with sharp surge in growth.

20 The particular interval of time is chosen merely to correspond with analogous data available in the union government budgets.

21 While this appears to be a comprehensive coverage, it extends beyond the scope of the working definition for the pay commissions. See also footnote 3 and opening paragraph under Section I.

22 The EFCB of the GoI covers Government Administration (GGA) and Departmental Commercial Undertakings (DCUs). Description of GGA appears to correspond closely with the definition of CGE as per the DGE&T. However, statutory corporations are not covered (perhaps rightly so) in the EFCB.

23 Crude estimates suggest that the number of military workers is more than 30 per cent of the number of civilian workers (as given in the budget). Thus, military constitutes roughly (or little less than) one fourth the size of the central government workforce.

24 We have used the cosumer price index for urban non-manual employees to adjust the series [RBI 2006].

25 Corresponding trend rates for the period, between 1988-89 and 2004-05, was 1.48 per cent, while between 1988-89 and 2005-06 this was 1.49 per cent per annum. Note that the rate is sensitive to the choice of terminal years as well as nature of estimates (whether actuals, revised (RE) or budget estimates (BE)). For example, this was 1.16 per cent per annum over the two-decade period from 1987-88 (RE) to 2006-07 (RE).

26 Corresponding trend rates for the period, between 1988-89 and 200405, was 3.50 per cent, while between 1988-89 and 2005-06 this was

3.42 per cent per annum. 27 In turn this is comparable to the trend rate immediately preceding the award of FPC (2.48 per cent per annum between 1988-89 and 1996-97).

28 Prior to the FPC, the annual increments at the mid point of the scales varied between 1.46 and 3.92 per cent (in the award of the Fourth Central Pay Commission, GoI, 1997).

29 The successive pay commissions have normally favoured an upward real revision often counter-balancing or compensating for differential in rates of growth.

30 The number of workers declined at the rate of 0.53 per cent per annum over 1988-89 and 2002-03. However, the rate was significantly steeper in the post-FPC period at 0.72 per cent per annum in comparison to

0.26 per cent per annum between 1988-89 and 1996-97.

31 Expenditure on pensions and retirement benefits for this sub-section are collated from Union Finance Accounts as EFCB presents a combined (capital) account for GGA and DCUs on gratuity payments and commuted value of pensions.

32 FPC and expert group reports also provide estimates for number of pensioners from large DCUs, namely, Railways, Posts and Telecom. One may also note that the extant number of pensioners from defence services exceeds the number of pensioners from civilian services.

33 An estimate, for the number of pensioners in the year 1996-97, is not easily available. Here we utilise the estimate available for 1994-95 and

Economic and Political Weekly August 4, 2007

34 Passivity ratio is the ratio of number of years (until death) in retirement to the number of years in active service.

Anand, Mukesh (2007): ‘Pensions’ in Kaushik Basu (ed), The Oxford Companion to Economics in India, Oxford University Press, New Delhi, India, pp 391-94.

Anand, Mukesh and Rajeev Ahuja (2004): ‘Government Pensions: Liability Estimates and Assumptions’, Economic and Political Weekly, Vol 39, No 25, pp 2569-76, June 19.

Anand, Mukesh, Amaresh Bagchi and Tapas K Sen (2004): ‘Fiscal Discipline at the State Level: Perverse Incentives and Paths to Reform’, Ch 3 in Edgardo M Favaro and Ashok Lahiri (eds),Fiscal Policies and Sustainable Growth in India, Oxford University Press, New Delhi, India, pp 80-101.

Census of India (1961, 1971, 1981, 1991, 2001): Final Population Totals, Registrar General and Census Commissioner, India.

Sage