“Performance-based Budgeting: A Review of Experience and Lessons”, Indian Journal of Social Work, 72 (4), October 2011, pp. 483-504.

Performance-based Budgeting A Review of Experience and Lessons

R. RAMAKUMAR

This article is a review of efforts to introduce performance-based budgeting in India. It attempts to discuss the context in which these new methods of budgeting have emerged and assess, with the help of published literature, their utility and impact. First, it deals with the global context in which the increasing concern with governmental efficiency has emerged. Second, it looks at the experiences of developed countries in the use of performance budgeting and results-based management (RBM). Third, it looks at the specific experience of India in the use of two major performance-based budgeting tools: ‘performance budget’ and ‘outcome budget’. Fourth, in the light of the evidence, the article suggests a sharp turn away from certain prevalent orthodoxies in public expenditure management towards adopting practises that help deepen democracy and transparency.

R. Ramakumar is Associate Professor, Centre for Development Studies, School of Social Sciences, Tata Institute of Social Sciences, Mumbai.

I

INTRODUCTION

There are two im por tant di men sions to the effectiveness of the state in pro -vid ing basic so cial ser vices to the cit i zens. First, a fea ture of so ci et ies that have sig nifi cantly im proved the con ditions of life of people is their high lev -els of pub lic spending on the so cial sec tors. As an il lus tra tion, com pare the case of India, which has one of the most back ward edu cational and health in -dicators1, with a few other coun tries with advanced health in di cators. On an av -erage, for the pe riod 2007–2011, India spent just about 1.4 per cent of its Gross Do mestic Product (GDP) on health. The cor re sponding share in the United States was 7.9 percent, in the United King dom was 7.8 per cent, in Cuba was 11 per cent and in Costa Rica was 7.1 percent. For the year 2008, India spent an av -erage of about 3 percent of its GDP on ed u cation. The correspond ing fig ure in

IJSW, 72(4), 483–504, October 2011

484 R. Ramakukar

the United States was 5.5 per cent, in the United King dom was 5.4 percent, in Cuba was 13.6 per cent and in Costa Rica was 6.3 per cent.

In other words, it is ex tremely im por tant to raise the lev els of pub lic spend ing on the so cial sec tor; ec ono met ric stud ies have shown that there is a high level of tem po ral as soci a tion between lev els of public spend ing and human development outcomes2. Of course, a one-to-one re la tion ship be -tween social spending and indicators of human development is difficult to make. How ever, higher lev els of pub lic ex pen di ture in the so cial sectors do denote a particular public policy stance, and public expenditure can be used as a good proxy to measure it.

Sec ond, there is an increasing concern over im prov ing the ef ficiency of public spending in the so cial sec tor. Such con cern is part of a grow ing strand of lit er a ture called as New Pub lic Man age ment (NPM), where gov ernment func tions are seen pri mar ily as man a ge rial. The NPM lit er a ture typ i cally ar -gues that im proving the effi ciency of gov ernm ent re quires it ‘to melt the fat”, “turn bureaucratic initiatives into entrepreneurial institutions, ready to kill off obso lete initiatives [and] will ing to do more with less…’ (Osborne and Gaebler, 1992, p. 23, em pha sis added). In other words, a rise in pub lic spend ing is seen as unneces sary, if gov ernm ents are able to function more ‘efficiently’. Fol lowing this lit era ture, a large num ber of new con cepts, such as performance budgeting, results-based management (RBM) and outcome budgeting, have emerged as dom i nant in the sphere of pub lic ad min is tra tion.

This ar ti cle dis cusses the sec ond di men sion listed above. First, it deals with the global con text in which the increas ing concern with gov ern men tal ef fi ciency has emerged. Sec ond, it looks at the ex pe ri ences of de vel oped coun tries in the use of per form ance bud geting and RBM. Third, it looks at the spe cific ex pe ri ence of India in the use of two major per for mance-based budgeting tools: ‘performance budget’ and ‘outcome budget’3. Fourth, in the light of the evi dence, the note suggests a sharp turn away from cer tain prevalent orthodoxies in public expenditure management towards adopt­ing prac tises that help deepen de mocracy and trans par ency.

II

THE EVOLUTION OF THE CONCEPT

There are clearly three phases in the emer gence and pop u lar ity of the con -cept of per form ance-based budget ing, as is seen today. The first phase is the phase of perfor mance bud get ing, which emerged in the United States in

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 485

the early part of the twen ti eth cen tury. The sec ond phase is the phase of New Pub lic Man age ment (NPM), where the focus was on cre at ing a mana -ge rial atm o sphere for the gov ern ment to per form like any other pri vate en -tity. The third phase is the phase where NPM ar gu ments came to be nested within new frame works, such as “good gov er nance” and “de centrali sa -tion”, as de fined by in terna tional or gani sations like the World Bank.

Phase 1: Early Forms of Performance Budgeting

The literature on public finance and public administration shows that de­mands for leg is la tive ac count abil ity of the state’s fi nances have been raised in vari ous coun tries from the 19th century it self. How ever, this account -abil ity was lim ited to the pre senta tion of a set of ac counts re flect ing the transactions of the gov ernm ent. In other words, the pro cess gener ally fol -lowed was of a fi nan cial or ap pro pria tion audit. The effectiveness of these ex pen di tures was never ana lysed in these doc u ments. Some authors have called this an input budgeting sys tem.

It is su premely im por tant to note that over the years, the idea of in tro -ducing performance-based assessment of public expenditure emerged pri-mar ily from deci sions to con trol or reduce govern ment spend ing. One of the first ef forts at as sess ing the per form ance of pub lic ex pen di ture was made in the United States in the 1930s. In 1938, Clar ence E. Ridley and Her bert A. Simon put to gether a vol ume titled Mea suring Mu nic i pal Ac tiv -ities (Ridley and Simon, 1938). This volume is con sidered as the first in a se ries of efforts that cul mi nated in the in troduc tion of per form ance bud get -ing in the US. The Hoo ver Com mis sion, chaired by Herbert Hoo ver and which subm it ted its re port in 1949, was an im portant milestone in this se -ries of ef forts ‘to as cer tain whether gov ernm ents were spend ing wisely’ (Premchand, 1993). The Hoo ver Com mission was asked ‘how ef fi ciently and economically an ap proved programme was being im ple mented, and how the same amount of work could be performed sat is factorily through other arrangements or through improved procedures at less cost’ (ibid., emphasis added).

The Hoo ver re port led to the in troduction of the per form ance bud get ing sys tem in the United States in the 1950s. This sys tem sought to change the ex ces sive em pha sis on in puts to bud get ary out puts. The per for mance bud -get ing sys tem also stressed on the mea sure ment of costs, better clas si fi ca -tion of government transactions, measurement of productivity in government departments and the evaluation of success of programmes

IJSW, 72(4), 483–504, October 2011

486 R. Ramakukar

after com ple tion. There is con sensus, how ever, that despite its many posi -tive features, per form ance bud get ing in the US was marked by poor im ple -mentation (Premchand, 1993). Efforts towards per for mance budgeting remained confined to just programme evaluation studies.

Phase 2: Performance Budgeting under New Public Management

Over time, with the sys tem of per form ance bud get ing fail ing to cut pub lic ex pen di tures ad e quately, focus shifted to wards cre at ing a ‘man a ge rial at -mo sphere’ that would be ‘atten tive to per for mance (Premchand, 1993). In other words, the focus was on putting in place a contin uous system of “good management practises, cost consciousness and cost responsiveness through out the manage ment chain”. As Premchand noted:4

…ascertaining the continued relevance of programmes in the changing economic environment was… emphasised. The major issue for the overall welfare of the community was (and is) whether some of the services provided could have been better performed outside the government sphere. Continuing pursuit of these themes contributed on the one hand to contracting out services on a competitive basis, and on the other to specifying an annual performance agreement with the agencies.

According to this new man age rial frame work, budgets were to be moulded into a ‘con tract for per for mance’ with peo ple. The contract ing out of pub lic ser vices to pri vate agen cies be came a major ob jec tive of this new frame work. Thus, in most coun tries where New Pub lic Manage ment (NPM) was adopted, it went hand-in-hand with the adop tion of in creas -ingly free-mar ket pol i cies aimed at down sizing the state. As Polidano (1999) noted:

In so far as we can generalise, two of the more commonly adopted elements of the NPM agenda are privatisation and downsizing (or retrenchment, as it is known in Africa). Such initiatives are part and parcel of the economic structural adjustment programmes which the majority of developing countries throughout the world have undertaken at some point in time; they are often the first stage of public sector reform (pp. 5-6).

According to another writer, the NPM had “be come a new pol icy fash -ion or fad, a perva sive zeit geist dif fused by in ter national bod ies, such as the Organisation for Economic Co-operation and Development (OECD), na tional gov ern ments, business schools, ar mies of consul tants, and re -search insti tutes” (Wright, 1994, cited in Suleiman, 2003). A large num ber of countries in the West adopted this frame work in the late-1970s, 1980s and 1990s. Some of the ini tia tives were the following:

1. United Kingdom: Financial Management Initiatives (FMI), 1982;

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 487

2. Canada: Increased Ministerial Authority and Accountability (IMAA), 1970s;

3. Australia: Financial Management Improvement Programme (FMIP) and Programme Man agem ent and Budget ing (PMB), 1980s;

4. Denmark: Budgeting for Productivity Improvement, 1983;

5. Sweden: Triennial Budgeting, late-1970s;

6. New Zea land: The Public Fi nance Act, 1989; and

7. United States: Gov ern ment Per for mance and Re sults Act (GPRA), 1993.

The evo lu tion of the NPM frame work in all the coun tries listed above evolved parallelly with the neo-lib eral ad vo cacy for pub lic ex pen di ture cuts. The aim for adopt ing the NPM was the main te nance of cur rent lev els of per for mance at re duced levels of spend ing, rather than im prove per for -mances concur rent to a rise in spend ing. The ‘fis cal con straint’ of the state was cited as the rea son why a rise in spend ing was not a fea sible op tion. The ex am ples of Aus tra lia, Denm ark and Sweden illustrate this point.

In Aus tra lia, the FMIP and PMB in the 1980s were in tro duced in a pe -riod when the gov ern ment was ac tively prac tising a “tril ogy” pol icy of “stabilising budget outlays, tax revenue and deficit as a percentage of the gross do mes tic prod uct” (Schick, 1990, p. 29). The ef fect, as Schick at trib -utes to of fi cial sources, was a “squeeze on re sources” (ibid.). On Den mark, Schick’s study notes the following:

Through most of the 1980s, the government has imposed a two percent cut on operating expenditures. The intention has been to force agencies to be more productive by squeezing the resources allocated to them. The 1985 Public Expenditure White Paper announced that future spending limits would be set in terms of desired improvements in productivity (Schick, 1990, p. 30).

In Swe den too, Tri ennial Bud get ing was a pol icy that went hand-in-hand with cuts in pub lic ex pen di ture. A study notes that the policy

…has been introduced following a long period, beginning in the late-1970s, during which annual across-the-board cuts of two percent (in real terms) were imposed on administrative expenditures. This ‘cheese slicer’, as the two per cent cut is often referred to, has been made more flexible in recent years by requiring agencies to save five per cent over a three-year period…The enforced cutbacks reflect the government’s determination to stabilise the size of the public expenditure relative to the gross national product (Schick, 1990, p. 30).

Avail able lit er a ture pro vides no ev i dence to the suc cess of NPM ini tia -tives. A study by two IMF schol ars concluded that the em pir i cal lit er ature on government-wide per for mance bud get ing “does not pro vide a basis for

IJSW, 72(4), 483–504, October 2011

488 R. Ramakukar

strong con clusions about the ef fi cacy of these sys tems” (Rob in son and Brum by, 2005, pp. 43-44). Re view ing the re sults from the United States, Gilmour and Lewis (2006) noted that there was “lit tle sys tem atic ev i dence thus far that perfor mance bud get ing, as it has been im plem ented in states and cities, has had a major im pact on budget ing de cisions” (p. 743). Reviewing results from across coun tries, Suleiman (2003) con cluded the fol low ing:

It appears clear that the NPM agenda is indeed problem-ridden. Its claim to be a new and modern solution that will cure all ills does not sand up against empirical and analytical tests, and it loses much of its grandeur when measured against historical comparisons. Even when we strip the NPM agenda to something approaching its bare essentials and view it as a mere attempt to align the public sector with the private sector, it remains a highly problematic concept (pp. 127-128).

Gilmour and Lewis (2006) also noted prac ti cal dif fi culties in the adop -tion of a per for mance-based bud geting sys tem. They pointed out that:

Performance budgeting is a troublesome enterprise because it is difficult to know how to use performance information. If a program performs poorly, does that mean that it should be cut because it is wasting money or increased so that it can do better?...A program that is performing poorly might perform better if given additional resources, whereas another very successful program may need no more than its current allocation…In practise, performance budgeting may reflect merit no more than traditional budgeting (p. 743–750).

Phase 3: Performance Budgeting under “Good Governance”

By the late-1990s, the failure of the Wash ing ton Consen sus was being in -creas ingly re cog nised at the global level. Along side, it was also becoming clear that the em pha sis on mana ge rial changes had failed to cre ate any pos -itive im pact on bud get ing meth ods. From the late-1990s on wards, in the global dis course on pol icy, per for mance budgeting came to be nested under the rhet o ric of “good gov ernance” by in sti tu tions like the World Bank. Good gov ernance did not just in clude account abil ity; it also in -cluded “de cen trali sa tion” and “civil so ci ety”. How ever, these terms had spe cific mean ings within the World Bank dis course.

Decen trali sa tion was an el e ment in the World Bank’s pol icy discourse from the early-1980s it self. The Bank’s po sition on de centrali sation evolved over time; be gin ning with a more ex plicit char ac terisation of de -cen trali sa tion as the flour ish of market-based ac tors and in stitu tions at the local level in the 1980s, the po sition was mod i fied into support ing de cen -trali sation within the larger frame work of “good gov er nance” in the 1990s. In the World De vel opment Report 1983, the World Bank argued that, de -cen trali sa tion “should be seen as part of a broader mar ket-sur ro gate

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 489

strat egy” (World Bank, 1983, p. 123), and that the aim was “re plac ing the bur den of cen tral ad min is tra tion by de cen tral ised mar ket forces” (ibid., p. 87). The Bank argued that “where gov ern ments are sim ply di vest ing them -selves of ac tiv i ties that they can not man age, the trans fer of own er ship [read, privatisation] it self con fers the autonomy that is needed to make decentralisation work.” The Bank also noted that:

Decentralizing is not solely a matter of involving a wider range of people in discharging the responsibilities of the public sector. Governments…can also make greater use of markets and prices, since they avoid the heavy administrative requirements of centralized planning controls. While greater reliance on markets may appear to carry risks, many governments have learned that their own interventions can easily misfire…The practical advantage of relying more on markets is that the public sector can then concentrate on improvements in those activities for which market solutions are inappropriate (World Bank, 1983, p. 5).

Thus, the World Bank in tegrated de cen tralisa tion into the policy frame -work set by neo-clas si cal eco nom ics. The Bank noted that “the par tic i pa -tion of local busi nesses can also play a cru cial role in de cen trali sa tion, shaping in cen tives at the local level” (ibid., p. 123). Thus, the concept of firm-level decentralisation was given more importance than political, or government-level, decentralisation (see Patnaik, 2000). At the firm-level (or any pro duc tion unit), de cen tralisa tion was pre mised up on the neo-clas -si cal as sump tion that when every firm in an econ omy maxi mises its re turns (or func tions ef fi ciently), the econ omy as a whole also reaches an ef fi cient equilib rium. The World Bank view of de centrali sa tion in volved firm-level decentralisation (to be attained through privatisation, deregulation and denationalisation) complementing government-level decentralisation so that every firm in the econ omy could at tain ef fi ciency. Ac tors in civil so ci -ety who are pri marily mar ket-based, es pe cially NGOs, were an im portant type among ‘firms’ en cour aged at the local level.5

By the 1990s, the po sition of the Bank evolved into one where the old NPM agenda, de cen trali sa tion and neo-lib eral eco nomic pol icy were har -moniously in terwo ven and nested within the larger slo gan of “good gov er -nance”. Shah (2006) noted these link ages from the stand point of re duc ing corruption in public service delivery:

The new public management (NPM) paradigms have clear implications for the study of localization and corruption as it argues for contractual arrangements in provision of public services. Such a contractual framework may encourage competitive service delivery through outsourcing, purchaser-provider split under a decentralized structure of governance. The NPM goals are harmonious with localization as greater accountability for results reinforces government

IJSW, 72(4), 483–504, October 2011

490 R. Ramakukar

accountability to citizens through voice and exit mechanisms. Conceptually, therefore, the NPM is expected to reduce opportunities for corruption…

The “good gov er nance” slo gan of the Bank did not rep re sent any sub -stan tive dif ference from its po si tion of the 1980s. As Jenkins (2001) noted, the operationalisation of the idea of good gov ernance included a package: “restructuring state bureaucracies, reforming legal systems, supporting democratic decentralisation and creating accountability-en­hancing civil societies.” In practice, nevertheless, “governance has come to be as so ci ated with in sti tu tions de signed to sup port mar ket-led de vel -op ment” (ibid.). Citing the work of Wil liams and Young (1994), Jenkins (2006, pp. 250–251) noted that the Bank’s en cour age ment of civil so ci ety in Africa was “nothing less than a backdoor at tempt to trans form Af rican so ciet ies from the ground up [to] per form the roles as signed to them in liberal political theory and neo-liberal economic policy.”6 This new con -cept of civil so ciety has been put to use for dif ferent purposes in the de -vel op ing world by the World Bank, aid or gani sations like the USAID and Western governments: to speed up neo-liberal reforms, topple democrati­cally elected governm ents and weaken gov ern ments with an avowed anti-imperial ist agenda. In deed, it was not in ci den tal that the fresh wave of pop ularity for the idea of ‘civil soci ety’ in the 1990s began with the promotion of the Solidarity movement in Poland.

In sum, if we look at the global ex pe rience, the idea of per for mance bud -get ing came to be used pri marily as an in strum ent to down size the state, con trol public spend ing and en cour age the role of pri vate play ers in the process of governance.

Such a critique of per form ance bud geting is not a re jec tion of the im por -tance of the quality of pub lic ex pen di ture. The ef fective translation of pub -lic spend ing into pub lic goods and ser vices is in deed an im por tant task for any pro gres sive state. In de pend ent of the lev els of spend ing, there is scope for improvement in the institutional mechanisms of public service deliv­ery. The at tempt in this sec tion, on the con trary, was to argue that under the neo-liberal policy frame work, and within the rhet o ric of NPM and “good gov er nance”, lev els of pub lic spend ing be came a ca sualty. Low and stag -nat ing, if not fall ing, lev els of pub lic spend ing, adversely affected the qual -ity of de liv ery of pub lic ser vices.

The In dian ex pe ri ence with per for mance bud gets, and later out come bud gets, needs to be dis cussed in this global con text. The au thor at tempts to do this in the next sec tion.

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 491

III

PERFORMANCE BUDGET AND OUTCOME BUDGET IN INDIA

Performance Budgets in India

The in troduc tion of per for mance bud gets in India was in spired by its in tro -duc tion in the United States.7 In 1953, Paul H. Appleby, a con sul tant with the Ford Foun da tion, was in vited by the then Fi nance Minis ter, C. D. Deshmukh, to study the possibilities of reforming public administration in India. Appleby, how ever, was not an en thusi ast of perfor mance bud gets in the United States: he noted in his re port that “the re cently in tro duced ‘per -for mance bud get’ in the United States has cer tain spe cial val ues, re veal ing what it is de signed to re veal, but not re veal ing what it is not designed to show” (Appleby, 1953, p. 35). Notwith stand ing Appleby’s ap pre hen sion, de mands to in troduce per for mance bud gets were made by mem bers of the Lok Sabha in 1954. In 1958-59, a fa vourable sug gestion came from the Es -timates Committee of the Lok Sabha; its report noted:

…the Performance-cum-Programme System of budgeting would be ideal for a proper appreciation of the schemes and outlays included in the budget, especially in the case of large scale developmental activities. Performance Budgeting should be the goal which should be reached gradually and by progressive stages without any serious budgeting dislocation. (cited in Madan and Mukerjee, 1989).

This recommendation was repeated by the Estimates Committee in 1960, which was accepted by the Fi nance Min istry in 1961. In 1964 and 1965, the Plan ning Com mission ini ti ated a set of pilot studies on per for -mance budgeting (Toye, 1981). According to the Planning Com mis sion, “the stage has reached when ap pro priate meth ods of Per for mance Bud get -ing should be evolved, so that these be come an in tegral part of the machin -ery for plan ning and su per vision over plan ful film ent” (cited in Madan and Mukerjee, 1989). In 1965, a Per for mance Bud get ing Unit was formed within the Com mission of fice. Toye (1981) gave spe cial em phasis to the tim ing of the Plan ning Com mission efforts: “it was at this time, when the politi cal lead er ship of the Con gress Party had become anxious and sceptical about the achieve ments of planning as it had oper ated dur ing the Nehru years, that the deci sion to introduce performance budgeting seems to have been taken” (p. 123).

The pilot stud ies of the Plan ning Com mis sion be came a da tabase that came to be analysed by a work ing group on per for mance bud gets of the

IJSW, 72(4), 483–504, October 2011

492 R. Ramakukar

first Ad minis trative Re forms Com mis sion (ARC) set up in 1965-66. In 1969-70, the first per for mance bud get was pub lished, ti tled Performance Budgets of Selected Organisations, 1968–1969. Over the years, some state governments also introduced performance budgets for specific depart­ments.

Eval uations of perfor mance bud gets after 1969 do not re port en cour ag -ing find ings. In a detailed eval uation, Toye (1981) re ported that while there was “steady prog ress” in purely quan ti ta tive terms, there was “lit tle cause for satisfaction” in qualitative terms. Toye argued that:

…the enthusiastic expectations which greeted the introduction of performance budgeting in India have been disappointed. Performance budgets have not been extended over the five-year planning horizon, new measures of the analysis of performance have not been developed and an integral system of planning and budgeting with ex post evaluation being used as feedback has not been constructed (p. 141).

Toye identi fied three major prob lems in In dian perfor mance bud gets. First, the expenditure classification used in the performance budgets were typ i cally “in stitu tional” and hence, does not allow one to ex am ine the dif -fer ent ac tiv i ties of the govern ment in a “sim pler and more funda men tal sense”. Sec ond, the in di ca tors of per for mance used were highly “in ade -quate” for any meaningful assessment of the government’s activities.8

Third, the form at for the prep a ration of per for mance bud gets was un chang -ing over the years, which showed “very lit tle ev i dence of ex per i ments or innovations”. Thus, performance budgets embodied “description”, and not “analysis”.

An im por tant prob lem with the way per for mance bud gets were at -tempted in India was its me chani cal ap pli ca tion, di vorced from local con -texts. As Toye (1981) noted, perfor mance bud gets were pre pared “in the spirit of rou tine doc u ment com pi la tion rather than of ad min is tra tive pi o -neer ing” (p. 126). In a broader re view across de vel op ing countries, Peter Dean noted that:

Performance budgeting was not tried, tested and proven as an operational system at the time developing countries adopted it. On the contrary, dissatisfaction was beginning to emerge in the USA about its usefulness. When it was exported to developing countries, these reservations were suppressed. Instead, exaggerated claims were made about what it would accomplish. Developing countries tried to fulfil these claims but they encountered a range of difficulties similar to those already experienced in USA (Dean, 1990, p. 16).

In sum, perfor mance bud gets in India did not lead to any improve ment in the ef fectiveness of pub lic spending. Nei ther did they meaning fully

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 493

pro vide in formation to bud get-mak ers on the di rec tions to adopt, nor did they pro vide in for ma tion to the pub lic on how well gov ern ment schemes are performing.

From Performance Budgets to Outcome Budgets

The fail ure of perfor mance bud gets to mean ing fully con trib ute to the bud -get ary pro cess was well-ap pre ci ated in the In dian of fi cial cir cles by the 2000s. The em pha sis, then, slowly shifted from perfor mance bud gets to ‘outcome budgets’.

An out come bud get is a doc u ment that links out lays of the gov ern ment to ‘out comes’ through the pro vision of in puts and the creation of out puts. There are, of course, di verse def i nitions for this term in the lit er a ture on public administration. According to one study, outcome budget is “any bud get that pres ents infor ma tion on what agen cies have done or ex pect to do with the money provided” (Curristine, 2005, pp. 130–131). More spe -cif i cally, it is de fined else where as “a form of bud get ing that re lates funds al lo cated to measur able re sults” (OECD, 2003, p. 3).

In any out come bud get, there are four terms of sig nif i cance: out lays, in -puts, out puts and out comes. Outlays are the fi nancial al loca tions made by the gov ernm ent in its budget on dif fer ent programmes. Inputs are the dif -ferent tangible and intangible provisions made by the government (and/or the ben e fi cia ries) in order to achieve the ob jec tives of these programmes. For example, scholar ships for stu dents or the in vest ment in the construc -tion of a school build ing in ed ucation-re lated schemes are con sid ered as in -puts of the gov ernm ent. Pri vate costs in curred by the ben e ficiaries and dif fer ent forms of vol un tary/free pro vi sions are also con sid ered as in puts. Out puts are the as sets and ser vices cre ated under the programme. For ex -am ple, the school build ing is an out put in the edu cation-related programme. How ever, the con struc tion of a school may not be suf fi cient to achieve the ob jec tive of con ceiv ing the scheme, which may have been to raise the en roll ment rates or re duce the dropout rates. This is where the dif -fer ence be tween an out put (the school build ing) and an out come (rise in en -rollm ent rates resulting from the con struc tion of a school) is of im portance. An out come is the re sult we want from a programme at the bot tom line, which fol lows the uti li sa tion of as sets and ser vices cre ated.

A snap shot of the major dif fer ences between per for mance bud gets and out come budgets is given in Table 1. An outcome bud get dif fers from a per -for mance bud get in that it fo cuses on how well the as sets and ser vices

IJSW, 72(4), 483–504, October 2011

494 R. Ramakukar

cre ated are uti lised, and not just cre ated. In our ex am ple of ed u ca tion-re lated schemes, an outcome bud get would docu ment the ex tent to which the con ­struc tion of a school build ing has been success ful in im prov ing, say, the en ­roll ment rate. It would thus ana lyse the ex tent to which the creation has been success ful in rais ing uti li sation. In cases where it is found to be fail ing, the out come budget would discuss the reasons for the poor uti li sation of the as ­sets. It would put for ward con crete suggestions for reforming organi sations and in sti tu tions so that uti li sation of as sets, and ser vices, could be im proved.

TABLE 1: Major Differences between Performance Budgets and Outcome Budgets

Performance Budgets Outcome Budgets

• Emphasis on outputs • Emphasis on outcomes and effectiveness

• Emphasis on creation of assets and • Emphasis on utilisation of assets and services services

• Seeks to optimise programmes • Seeks to improve organisations for better spread and reach of programmes

• Emphasis on processes, compliance • Emphasis on results, performance and and management control management improvement

Source: collated from different sources.

While the term out come budget earns its name from the em pha sis on out comes rather than out puts, the con cept is used under dif ferent names in dif fer ent coun tries; the re lated terms in clude RBM, en tre pre neur ial bud ­get ing and mis sion budgeting.

The In dian govern ment first an nounced its plan to launch out come bud ­gets in the Fi nance Min is ter’s bud get speech for 2005-2006. Fol low ing this an nounce ment, Prime Min is ter Manmohan Singh wrote to all min is ­ters in 2005 stress ing the im por tance of “converting fi nan cial out lays into phys i cal out comes, with fixed quar terly mea sur able and monitorable tar ­gets, to improve the quality of implementation of development programmes” (cited in Gupta, 2010). To coor di nate the prep a ra tion of out ­come budgets, the Plan ning Com mis sion cre ated a new di vi sion called the ‘Programme Out come and Response Mon i toring Di vision’ in 2005. A core group was set up in 2005 itself with of fi cers from the Plan ning Com mis ­sion and the Min is try of Fi nance to issue uni form guide lines for the prepa ­ra tion of out come budgets. The first out come budget for the year

IJSW, 72(4), 483-504, October 2011

Performance-based Budgeting 495

2005–2006 was re leased in 2006; this was fol lowed by sim i lar budgets for 2006–2007, 2007–2008, 2008–2009, 2009–2010 and 2010–2011. The ear -lier per form ance bud gets were dis con tin ued and merged into the prepa ra -tion of out come bud gets. In the foreword writ ten for the outcome bud get of 2005–2006, the Fi nance Min ister noted the im por tant steps in the pro cess:

• Outcomes should be specifically defined in measurable and monitorable terms, and intermediate outputs should also be defined wherever required;

• Standardising unit cost of delivery; • Benchmarking the standards/quality of outcomes and services; • Capacity building for requisite efficiency at all levels, in terms of

equipment, technology, knowledge and skills; • Ensuring flow of right amount of money at the right time to the right

level, with neither delay nor ‘parking’ of funds; • Effective monitoring and evaluation systems; and • Involvement of the community/target groups/recipients of the

service, with easy access and feedback systems.

There have not been many evalu ations of out come budgets of the cen tral gov ern ment, but avail able stud ies note lit tle dif fer ence be tween the ear lier per for mance bud gets and out come bud gets. It would appear that the rea -sons for the fail ure of per for mance bud gets were, clearly, not ap pre ci ated and the same er rors are al lowed to be re peated. Even staunch propo nents of outcome budgets have expressed disappointment; Premchand (2006) noted that “the Outcome Budget (OB) as is for mu lated now is noth ing more than the per for mance bud get” (p. 1). Mathur (2010) re viewed the first six out come bud gets of the cen tral gov ern ment and con cluded thus:

A study of these budgets does not provide any clue as to how various schemes and programmes are functioning and the intended outcomes are being achieved. The enthusiasm with which Outcome Budget was launched in February 2005 seems to have petered out, as in the budget speech of the Finance Minister for 2008–2009, 2009–2010 and 2010–2011 there is not a single line making a mention of the Outcome Budget as a tool for monitoring performance of various schemes and development programmes…

...the Outcome Budgets of each ministry have become a huge mass of incoherent data and figures, having no relation with outcome of development schemes with which the public is concerned…It is obvious that Outcome Budgets of various ministries are not serving any purpose and have become an exercise in futility. (Mathur, 2010).

Gupta (2010) ar gued that the failure of out come budgets in India had many rea sons. First, gov ern ment of fi cials did not have any spe cific in cen tive to pre pare out come bud gets in any mean ingful way. Sec ond, no min is try or

IJSW, 72(4), 483–504, October 2011

496 R. Ramakukar

de partm ent had the req ui site skills or tech niques required to prepare out -come bud gets. Ac cord ing to him, due to these and many other rea sons,

the [outcome budget] has failed. This failure has happened because the assumptions of the theory of change underlying the [outcome budgets] have not been satisfied. It is business as usual, with no visible improvement in the effectiveness of the GoI’s spending, and with no positive change even in the mindset of those who manage the GoI’s spending. (2010, p. 571)

A com mon thread that runs through the three cri tiques pre sented above is an as sump tion that one al ready has ro bust meth od olo gies on the table to pre pare meaning ful out come budgets. In this sense, all the three critiques above miss the point. In fact, one of the most in trigu ing el e ments in the preparation of an outcome budget remains its methodology. The reports of the Gov ern ment of India are more forthright in ad mit ting this la cuna than the three cri tiques listed above. Ac cord ing to Montek Singh Ahluwalia, the Deputy Chairman of the Planning Commission:

The exercise of outcome budget of manipulating existing data did not get the desired results. It is easy to evaluate whether investment made in a steel plant has delivered. But measurement of outcomes in health and education are difficult and complex.9

In the draft ap proach paper for the elev enth five year plan, the Plan ning Commission noted that “there are internal difficulties” in the preparation of out come bud gets. The ap proach paper also gave two exam ples to illus trate these “difficulties”, which were primarily methodological:

In the case of education for example, starting from expenditures as a first step, one can track intermediate outcomes such as the construction of school facilities, the filling of vacancies and training of teachers, success in enrolment and reductions in drop out rates but the final outcome is really the quality of education provided. All the intermediate steps are relevant and also well worth monitoring, but the ultimate test of the strategy must be defined in terms of final outcomes. Often this can only be determined over a period of time. At times the causal chain may involve factors outside the program itself. This is evident, for example, in the case of health where improvements in health status are often the result not of interventions in the area of curative health but in the provision of clean drinking water and sanitation (Planning Commission, 2006, pp. 94-95).

These methodological difficulties in preparing an outcome budget should be ob vi ous. As long as there is no one-to-one re la tion ship be tween the in puts and out comes (that is, as long as ex trane ous fac tors in de pend -ently affect outcomes), the methodological difficulties in preparing out­come budgets would re main. Such a method ologi cal prob lem can not be transcended by assum ing away the ex tra neous fac tors or by try ing to ‘con -trol’ for them. These dif fi cul ties need to be ac cepted and al ter na tive ways of evaluating programmes and schemes need to be explored.

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 497

IV

LESSONS AND ALTERNTIVE DIRECTIONS

To con clude, first, the fail ure of out come bud gets at the central level in India brings to fore the per ils of me chan i cally adopt ing meth ods and tech -niques that are de vel oped out of his tori cal con texts and local re ali ties. This, pre cisely, was the prob lem with per form ance bud gets of the 1970s, and they re main the prob lem with out come budgets in the 2000s.

Second, performance-based assessments of budgets in developing coun tries, which do not spend ade quately in pro vid ing their cit i zens with basic needs, should be delinked from efforts to re duce ex pen di ture. Im -proving the efficiency of public expenditure is a necessary measure inde-pend ent of the lev els of spend ing. A major chal lenge in the prep ara tion and implementation of outcome budgets in developing countries, then, is to res cue it from the con ser va tive “sound fi nance” ar gu ments of neo-lib eral economists and international financial agencies.

Third, any method that mechani cally links per for mances with bud get al­locations sits uncomfortably with a rights-based approach to development. When cer tain basic needs of the peo ple are re cog nised as rights, there is lit -tle room for any cost-bene fit analy sis on the mat ter. Even if there is a ‘loss’ in the pro cess of achievem ent of an out come re lated to basic needs, the government cannot withdraw from its responsibility of provision in the name of inefficiency.

Fi nally, an im por tant prob lem with the methods and tech niques used in out come budget ing is that they have re mained in su lated from poli tics. As long as out comes are medi ated by po liti cal fac tors, outcome bud gets can­not as sume pol i tics away and take a tech no cratic stance. It needs to be re -cognised that budget making cannot be envisaged as independent of pol itics. The au thor would argue that any dis cussion on im provem ents or alternatives to outcome budgeting has to begin with the appreciation of politics in budget making in a fundamentally democratic manner.

The au thor indi cates three major forms of de par ture from the prev a lent or tho dox ies, all three of which in volve an ap pre ci a tion of pol i tics in pub lic ad min is tration. First, the links be tween the nature of pub lic service de liv­ery and the re forms re quired in the ad min is tra tive struc ture of the state has been an im por tant theme in the In dian lit era ture from the 1950s on wards. The Gorwala Com mit tee of 1950, the Appleby re ports of 1953 and 1956,

IJSW, 72(4), 483–504, October 2011

498 R. Ramakukar

the Santhanam Committee of 1962, the Administrative Reforms Commis­sion of 1966 and the Sarkaria Com mis sion of 1983 were all dif fer ent en -qui ries into the reforms re quired in Indian pub lic ad min istra tion to make it more ef fi cient and ef fec tive. How ever, very few of the rec om men da tions of these com mittees have been im ple mented till date. In stead, dis pro por -tionate em pha sis is being placed on per form ance-based budget ing meth ods to im prove pub lic ser vice de liv ery. The dan ger here is the side-lin ing of fun da men tal changes in fa vour of su per fi cial changes.

Sec ond, it is im por tant to re cog nise the need and im portance of de cen -tralisation – financial and political – in the efficient delivery of public ser­vices. The au thor shall try to il lus trate the im por tance of decen trali sa tion in pub lic ser vice de liv ery by briefly high light ing the ex pe rience of one state – Kerala.

De cen trali sa tion of gov ern ment (called as the Peo ple’s Plan Campaign, or PPC) was a major po lit i cal and ad min is tra tive ini tiative of the Gov ern -ment of Kerala from 1996 on wards. The PPC in volved a his toric de ci sion to de volve 35 to 40 per cent of the plan funds to the local gov ernm ents. The local gov ern ments were to draw up plans and pro jects for the de velopment of their re gions using the plan funds de volved. The par tici pa tion of all sec -tions of peo ple in the grama sabhas (village assem blies) and other fo rums of the plan ning process trans formed this planning pro cess from a mere ad -ministrative exercise to a mass movement for decentralisation.

Social scientists have given inadequate attention to the impact of post-1996 decentralisation in Kerala on local-level corrup tion and nep o -tism. In one anal y sis, T. M. Thomas Isaac ar gues that de cen trali sa tion “had a positive and significant impact on the reduction of corruption, nepotism and leak age” in pro ject im plem en tation (see de tailed table in Ap pen dix 1). Higher levels of transparency in decentralisation made corruption more vis i ble, which, in fact, was in strum en tal in re duc ing corrup tion. The les son from Kerala is that decentralisation is primarily a political transfer of power and such re dis tri bu tion of po lit i cal power can have sig nif i cant im -pacts on the na ture of pub lic service de livery and its ef fi ciency. However, pro po nents of performance-based budgeting have had nothing to offer on such fundamental changes required.

Finally, it is im portant to un derline the im por tance of evalu ation by the peo ple them selves. The au thor is re ferring to so cial audits here. A so cial audit is a pro cess where the ben e fi cia ries of a programme come to gether and exchange ideas, views and opin ions on all the in formation re gard ing

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 499

the programme from the standpoint of their per sonal ex pe rience. Ulti ­mately, the per for mance of a programme has to be mea sured in terms of how peo ple ex peri ence and judge the programme and its ben e fits. A social audit is also, thus, a po lit i cal process, which again has no place in a typical performance-based budgeting system.

In conclusion, alternative directions that performance-based budgeting should take has to be in terms of ap pre ci ating the cen tral ity of pol i tics in public administration. Unfortunately, changes in these directions do not ap pear to be forthcom ing under the cur rent dom i nant para digms of tech no ­cratic gov er nance. To con test and re write such par a digms is a major po liti ­cal challenge that the contemporary world faces.

APPENDIX

APPENDIX 1: Impact of decentralisation on different components of local corruption in Kerala, 1997-2001

Type of Corruption Impact of Rationale Decentralisation

Nepotism in beneficiary Drastically Introduction of transparent and selection reduced participatory beneficiary selection

procedure

Corruption at the Eliminated Administrative sanction made administrative sanction automatic with the approval of the stage of public works plan

Corruption at the Virtually Not a single allegation of corruption technical sanction stage eliminated raised against expert committees despite of public works controversies that surrounded them

Corruption in the award Significantly Beneficiary committees promoted of works to the reduced instead of contractors contractors

Corruption during Reduced Only less than 25 per cent of the actual implementation beneficiary committees were benami

committees. Nearly 25 per cent of the beneficiary committees were exemplary and raised substantial additional resources.

Frauds in the purchase Reduced Better monitoring and public visibility of materials than in the past

Source: Thomas Isaac (2007).

IJSW, 72(4), 483-504, October 2011

Type of Corruption

Nepotism in beneficiary selection

Corruption at the administrative sanction stage of public works

Corruption at the technical sanction stage of public works

Corruption in the award of works to the contractors

Corruption during actual implementation

Frauds in the purchase of materials

Impact of Decentralisation

Drastically reduced

Eliminated

Virtually eliminated

Significantly reduced

Reduced

Reduced

Rationale

Introduction of transparent and participatory beneficiary selection procedure

Administrative sanction made automatic with the approval of the plan

Not a single allegation of corruption raised against expert committees despite controversies that surrounded them

Beneficiary committees promoted instead of contractors

Only less than 25 per cent of the beneficiary committees were benami committees. Nearly 25 per cent of the beneficiary committees were exemplary and raised substantial additional resources.

Better monitoring and public visibility than in the past

500 R. Ramakukar

NOTES

1. All data used here are from http://data.worldbank.org

2. In a recent study using panel data from 120 developing countries between 1975 and 2000, Baldacci and others (2004) argued that public spending in education and health had “a positive and significant direct impact on the accumulation of education and health capital” (p. 28). They argued that “an increase in education spending of 1 percentage point of GDP is associated with 3 more years of schooling on average…Similarly, an increase in health spending of 1 percentage point of GDP is associated with an increase of 0.6 percentage points in the under-5 child survival rate…”

(ibid.). These pos i tive im pacts were stron gest in the case of low-in come coun tries and countries from Af rica. In their widely-cited study of 22 coun tries, Anand and Ravallion (1993) ar gued that the ef fect on life ex pec tancy rates of an in crease in pub lic spend ing on health was twice as sig nifi cant as an in crease in per ca pita in come. They con cluded that “cer tain com po nents of pub lic spend ing can mat ter greatly in en hanc ing human de vel op ment in poor coun tries, and that they mat ter quite in de pend ently of what they do or don’t de liver in terms of re duced in come pov erty” (p. 147). Chakraborty’s (2003) analy sis for selected developed and developing countries showed that “pub lic spend ing on ed u ca tion and health has a stron ger im pact on human development than the growth of per capita income”.

3. Throughout this paper, I have dis tin guished be tween the terms “per formance-based bud get ing”, which is a ge neric term, and “per for mance bud get”, which is a spe cific applica tion of the concept.

4. http://unpan1.un.org/intradoc/groups/public/documents/APCITY/UNPAN014322.pdf

5. The idea of firm-level decentralisation is theoretically misplaced. Prabhat Patnaik has argued that firm-level and government-level decentralisation stand contradic tory to each other as firm-level-decentralisation actually “emasculates decentralised planning and makes gov ern ment-level de cen trali sa tion vac u ous” (Patnaik, 2000, p. 10). According to him, “decentralised decision-making, even when informed by individual ‘rationality’, is fundamentally ‘irrational’ because the units are in reality not separable from one an other” (ibid., p. 2). Every eco nomic de ci sion taken by a firm af fects other firms as well. Due to this inter-re lated na ture of firm-level ac tions and outcomes, the final out come would be ir ra tio nal for all the firms al though each firm would have acted ra tio nally. This fal lacy of firm-level de cen trali sation can be corrected, (in other words, a more ra tio nal out come for all firms to gether en sured) if every firm col luded with every other or if the state in ter vened through eco nomic plan ning. The former being a remote possibility, the latter emerges as the only possible solution.

6. Singh (2005) notes that in the 1990s, the USAID funded 60 “local gov er nance” ini tia tives while the UNDP funded more than 250 “de cen trali sa tion” pro jects. Ac cord ing to him, “in the name of de cen trali sa tion and local self-gov er nance, es sen tial de vel op men tal tasks and so cial re spon si bil i ties of the state are being handed over to cash-starved, non-trans par ent, un ac count able NGOs and local bod ies with out examining their performance and capacity to deliver…[T]he welfare functions of the na tion-state are…being weak ened from below in the guise of decentralisation” (p. 121).

7. The United Na tions too pub lished a man ual for Programme and Per for mance Bud get ing in 1965 it self (Dean, 1990). As Toye (1981) noted in jest about the in tro duc tion of per for mance bud gets in India: “…Govern ment of India tends to seize with alacrity on foreign administrative innovations…” (p. 122). Dean (1990),

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 501

how ever, cau tions against the read ing that out sid ers thrust per for mance bud get ing on de vel op ing coun tries. He notes that the rul ers in these coun tries also en thu si as ti cally ac cepted the model, just as experimenters in the United States did.

8. According to Toye, “…the com bi na tion of fi nan cial and phys i cal data on pro ject’s tar gets and ac tual achieve ments do not allow us to reach safe con clu sions about the per for mance of any par tic u lar pro ject.” (1981, p. 125).

9. “Outcome budget serves no pur pose, says Montek”, The Eco nomic Times, 13 Oc to ber 2009, avail able at: http://ar ticles.economictimes.indiatimes.com/2009-10-13 /news/28414324_1_outcome-budget-monetary-policy-stimulus-package

REFERENCES

Anand, S. and : Human Development in Poor Countries: On the Role of Private Ravallion, M. Incomes and Public Services, Journal of Economic 1993 Perspectives, 7, Winter.

Appleby, P.H. : Public Administration in India: Report of a Survey, New Delhi: 1953 Manager of Publications.

Baldacci, E., : Social Spending, Human Capital, and Growth in Developing Clements, B., Countries: Implications for Achieving the MDGs, IMF Working Gupta, S. and Paper No. WP/04/217, Washington, D.C.: International Cui, Q. Monetary Fund. 2004

Chakraborty, L.S. : Public Expenditure and Human Development: An Empirical 2003 Investigation, Working Paper, New Delhi: National Institute of

Public Finance and Policy.

Curristine, T. : Government Performance: Lessons and Challenges, OECD 2005 Journal on Budgeting, 5(1), 127–151.

Dean, P.N. : Government Budgeting in Developing Countries, London: 1990 Routledge.

Gilmour, J. and : Assessing Performance Assessment for Budgeting: The Lewis, D.E. Influence of Politics, Performance, and Program Size in FY 2006 2005, Journal of Public Administration Research and Theory,

16 (2), 169–186.

Gupta, A.P. : Evaluation of Governance: A Study of the Government of 2010 India’s Outcome Budget, Journal of Development Effectiveness,

2(4), 566–573.

India: Planning : Towards Faster and More Inclusive Growth: An Approach to Commission the 11th Five Year Plan (2007-2012), New Delhi. 2006

Jenkins, R. : The Emergence of the Governance Agenda: Sovereignty, 2001 Neo-Liberal Bias and the Politics of International Development.

In V. Desai and R. Potter (Eds.), The Companion to Development Studies, New York : Oxford University Press, 485–488.

IJSW, 72(4), 483–504, October 2011

502 R. Ramakukar

2006 : Mistaking ‘Governance’ for ‘Politics’: Foreign Aid, Democracy and the Construction of Civil Society. In S. Kaviraj and S. Khilnani (Eds.), Civil Society: History and Possibilities, Cambridge : Cambridge University Press, 250–268.

Madan, G.R. and : Economic Problems of Modern India: Problems of Mukerjee, R. Development, New Delhi : Allied Publishers. 1989

Mathur, B.P. : Outcome Budget: An Exercise in Futility—Redesign for 2010 Delivering Performance. In Alternative Economic Survey,

India: Two Decades of Neoliberalism, New Delhi : Daanish Books.

Organization for : Public Sector Modernization: Governing for Performance, Economic Policy Brief, October, Washington, D.C. Co-operation and Developnment 2003

Osborne, D. and : Reinventing Government, Reading : Addison-Wesley Publishing Gaebler, T. Company. 1992

Patnaik, P. : Alternative Paradigms of Economic Decentralisation. Paper 2000 presented at the International Conference on Democratic

Decentralisation, organised by the Kerala State Planning Board, Thiruvananthapuram, May 23-27.

Polidano, C. : The New Public Management in Developing Countries, IDPM 1999 Public Policy and Management Working Paper No. 13,

Manchester: Institute for Development Policy and Management, University of Manchester.

Premchand, A. : Public Expenditure Management, Washington, D.C. : 1993 International Monetary Fund.

2006 : False Dawn on Budget Front, (Unpublished), Hyderabad.

Ridley, C.E. and : Measuring Municipal Activities: A Survey of Suggested Criteria Simon, H.A. and Reporting Forms for Appraising Administration, Chicago : 1938 International City Managers’ Association.

Robinson, M. and : Does Performance Budgeting Work? An Analytical Review of Brumby, J. the Empirical Literature, IMF Working Paper WP/05/210, 2005 Washington, D.C.: International Monetary Fund.

Schick, A. : Budgeting for Results: Recent Developments in Five 1990 Industrialized Countries, Public Administration Review, 50(1),

26–34.

Shah, A. : Corruption and Decentralized Public Governance, World Bank 2006 Policy Research Working Paper No. 3824, Washington DC.:

World Bank.

Singh, K. : Questioning Globalisation, New Delhi : Madhyam Books. 2005

IJSW, 72(4), 483–504, October 2011

Performance-based Budgeting 503

Suleiman, E. : Dismantling Democratic States, Princeton: Princeton 2003 University Press.

Thomas Isaac, T.M. : Kerala’s People’s Plan Campaign, 1996–2001: A Critical 2007 Assessment, (Unpublished), Trivandrum.

Toye, J. : Public Expenditure Reforms in India and Malaysia, Development 1981 and Change, 12, 121–144.

World Bank, The : World Development Report 1983, Washington DC. 1983

IJSW, 72(4), 483–504, October 2011