Recently, somewhat quite suddenly, there has surfaced a threat to the autonomy of the Indian Statistical System (ISS), attributable to various actions on the part of official agencies, particularly after the economic situation began deteriorating. Against this backdrop, we study the working of ISS and also take this opportunity to address the wider issues faced by the statistical institutions and the system as a whole.
Evolution of the ISS
In this section, we look at the evolution of the ISS through the lens of its varied institutions.
CSO for coordination: World WarII gave fillip to the development of statistics for the government’s requirements and in 1945, the idea of forming “a central statistical office for the coordination” at the central and provincial levels was conceived. With P C Mahalanobis at the helm of affairs, right when planning commenced in 1951, the Central Statistical Organisation (CSO) was set up and
it was to bring about coordination of statistical activities amongst various statistical agencies in the central government and of statistical bureaus of state governments, which were set up for similar coordination of activities of statistical agencies at the state level. (NSC 2001: 433)
With the complexity of the semi-federal set-up, the issue of coordination between various levels of government agencies became the most crucial aspect of the statistical system and the CSO became the focal point for coordination.
NSSO for sample studies: Simultaneously, sensing the large data gaps for computing national income and at the instance of the National Income Committee and the then Prime Minister Jawaharlal Nehru, Mahalanobis devised a scheme of National Sample Survey (NSS), which would apply sampling techniques to fill the data gaps. It was approved by the government in 1950. Operationally, the sample survey work was divided between the NSS at the government level and the Indian Statistical Institute (ISI), Calcutta, set up in 1931, which had been carrying on large-scale sample surveys since the mid-1930s.
While fieldwork was organised by the NSS office, the ISI did the survey design, data processing and report writing. But, over time, such a division of responsibilities was said to have “caused extensive delays and created many problems in the organisation of the survey” (Rao 2012). Around the time of difficult economic situations in 1966 or thereabout, the Indira Gandhi government, facing tremendous pressures on poverty estimations, appointed a high-level committee under the then Cabinet Secretary, B Sivaraman. The committee recommended the setting up of the National Sample Survey Organisation (NSSO) “with a view to bringing the different stages of data collection, processing and publication under the same umbrella” (Rao 2012) and the NSSO was thus born on 5 March 1970.
The statistical commission: But the vast array of statistical infrastructure created, both at the centralised and decentralised levels, began to crumble towards the end of the 1980s, just alongside the crumbling of the overall governance system. “Government statistics are the first victims of this widespread failure of the system of government administration,” so said Vidwans, a member of the Rangarajan Commission (2002: 3824). Economic reforms of the 1990s, along with rapid structural changes taking place in the economy, made the traditional data sources further unequal to the demands of growing and complex statistical requirements.
The deterioration in the statistical system compelled the government to appoint the National Statistical Commission (NSC) in January 2000, with C Rangarajan as the chairman. The Rangarajan Commission was unique as never in the past had there been such a comprehensive report covering all aspects of the statistical system—institutional and administrative structure, the extent of data gaps in every ministry and department of the central government as well as at the states levels, and recommendations to improve them, including the suggestions for 10 census studies and about 60 types of sample studies. A path-breaking recommendation related to the creation of a permanent and statutory National Commission on Statistics (NCS) by an Act of Parliament, to be independent of the government and responsible only to Parliament. And amongst the multifarious tasks assigned to it, the following stands out:
TheNCS will ensure that the production of statistics and their release are free from government influence, by designating appropriate statistical institutions or functionaries to be solely and independently responsible for these functions. (NSC 2001: 457)
As an apex authority, the NCS was to operate through the statistics wing of the Ministry of Statistics and Programme Implementation (MOSPI). The latter was to be restructured as the National Statistical Office (NSO). The three key offices of the statistics wing were to be renamed as Central Statistics Office (CSO), National Sample Survey Office (NSSO), and Data Storage and Dissemination Office (DSDO), mainly replacing in each the word “Organisation” by “Office.” Also, a new consultancy wing was to be created.
On the basis of the above recommendations of the Rangarajan Commission, the government decided to set up the NSC, but not as a statutory body. Instead, it was done in June 2005 through a government resolution, with part-time chairman and members. Even so, the basic functions envisaged for the NSC were retained. Also, alongside, a decision was taken to bring the CSO and the NSSO under one umbrella—namely, the NSO—similar to the Rangarajan Commission recommendations.
In reality, such an NSO had remained only on paper. It is said that it was kept in abeyance for unknown reasons (Srivastava 2019b). As a result, the aforementioned proposal of an extra hierarchical set-up did not become operational and the system moved on only with the CSO and NSSO, pursuing their allotted responsibilities rather solely and independently. It seems to us that there was a strong justifiable reason for the proposal to remain in abeyance, that was, the autonomy exercised by the two institutions had a powerful logic, which merging them would have negated.
Thus, these two institutions retained their operational independence in their respective spheres. The CSO, for instance, compiling and releasing the National Accounts Statistics (NAS) series, index of industrial production (IIP), and consumer price indices (CPI), and conducting the economic censuses; as also coordinating for its statistical needs with central government departments and ministries as well as with the state government and union territory administrations, including the strengthening of existing institutional mechanisms; and the NSSO undertaking a series of sample studies, preparing their release calendars and actually releasing their field results. With a view to providing professional and technical inputs, the NSSO has been appointing, with the permission of the NSC, working groups for every round of its surveys with representation for domain expertise in all subjects covered in a round, and also representation to the concerned ministries/departments administering the subjects surveyed. Likewise, the CSO has a regular advisory committee with wide terms of reference concerning the NAS compilation, its methodology, data collection systems, studies for its improvement, etc.
A major task performed by the NSC in the above respect concerns the exercising of the functions of the erstwhile Governing Council of the NSSO. These functions performed by the NSC covered the decisions on the survey subjects, formulating sampling methodology and overseeing the processing of data and granting approval for the release of survey reports, even unit-level data, by the NSSO.
Merger of CSO and NSSO: Finally, in a very recent order dated 23 May 2019, MOSPI has undertaken a major step of what it called “restructuring” of the ISS. By restructuring, the order meant that “the CSO and NSSO to be merged into NSO.” In other words, the statistics wing of the MOSPI now comprises of NSO with CSO and NSSO merged as forming its constituent parts and the whole structure has been made “an integral part of the main ministry.” There is not an iota of mention of the NSC in this exercise.
The whole “restructuring” exercise has thus opened up a number of questions, which we take up for a critical review in the next section. For the present, we may conclude this section by presenting an overview of the current restructured statistical system:
(i) It is the ministry at the top as always, but its statistics wing now consists of NSO with CSO and NSSO as its constituent units and all of these forming an integral part of the ministry.
(ii) The NSO is headed by the chief statistician of India (CSI)-cum-secretary (statistics and programme implementation) with various divisions reporting to them through their directors general.
(iii) All divisions of the CSO, NSSO and the administrative wing exist as divisions of the MOSPI. Even the field operations division (FOD) of NSSO forms a subordinate office of the MOSPI. The other division of NSSO, namely, data processing division, is renamed as the data quality assurance division (DQAD) with enhanced role of improving the survey data as well as the data from non-survey sources like the economic census and administrative statistics.
In the press releases on the national income estimation since 31 May 2019 or in the latestPeriodic Labour Force Survey (PLFS) 2017–18, which was finally released in May 2019, new institutional nomenclatures have been given effect to. These apparently represent a tectonic shift implying a drastic change from being formally independent of the political establishment (ministry) to being part of it. This is so not only in the sphere of administrative tasks, which has always been the case, but also in substantive professional and technical works of statistical estimations. The implications of these changes for the autonomy of the statistical system are indeed disturbing and the issue is discussed in the next section.
‘Autonomy’ of ISS
In the recent period, two crucial issues have arisen in this respect. First, there have been concrete instances of political interference in the functioning of the system. Second, the structured edifice of the system, the way decisions are taken to build the edifice, appear to show less respect for the essential issue of the autonomy of the institutions.
Encroached autonomy: Recently, three distinct cases of such encroachment have been widely exposed in the media.
The first one concerns the National Institution for Transforming India (NITI) Aayog’s persistent interest shown in the release of the back series of national income by the CSO because it had implications for the National Democratic Alliance (NDA) government’s growth performancevis-à-vis that of the previous United Progressive Alliance (UPA) government. All technical works done, either by the CSO initially towards the end of 2015 or subsequently by an NSC sub-committee in July 2018, were showing the average growth rates of the UPA period somewhat higher than that of the first four years of the NDA government, which the NITI Aayog was not prepared to accept (Iyer 2018). And the final release of the CSO’s back series was done in November 2018 in an unusual way. It was done in the NITI Aayog by its vice chairman, Rajiv Kumar, along with CSI-cum-secretary Pravin Srivastava. This final release reduced the growth rates for all UPA years with drastic reductions for some years, which we suspect to have happened because of the arbitrary interventions by the powers that be (Rajakumar 2018)—a blatant attempt to alter the statistical results in the NDA government’s favour. This was also an obvious case of compromising with the independence of CSO.
The second instance concerns the sudden merger of the CSO and the NSSO into one entity called NSO, as explained earlier. The justification for the merger provided in terms of avoiding duplication between the two institutions (Srivastava 2019a) appears far-fetched. The nature of works done by the two institutions is by and large independent—one, what we may call, doing desk work, while the other is fieldwork. Within this broad separation of functions, the CSO shares work with the NSSO where fieldwork is involved, such as in conducting the annual survey of industries (ASI) and in constructing the CPI for the agricultural and rural labourers; and the urban and rural consumers. In such survey-based tasks, the NSSO’s FOD carries out fieldwork and the industrial statistics wing of the CSO processes the data and publishes the results. The division of labour is so clear-cut, that there is no scope for duplication getting involved. No doubt suitable coordination has to be ensured.
Also, the inference that the new arrangement would provide greater freedom to the CSO and the NSSO in their operations after the creation of another tier in the form of NSO, appears unrealistic and indefensible. Srivastava is reported to have said that “the CSO and NSSO worked as subordinate offices of the MOSPI and officials working there felt that they had no authority.” This was when they were professionally independent even if in an unwritten way. Now, the statistics wing has taken formal control over them in the form of NSO and therefore, they will have to operate under another formal tier over them, both administratively and professionally. What appears certain to take place is what Srivastava admitted that the merger will help meet, that is “the requirement of the statistical system as a lack of control on these two bodies was one challenge presently.” This is a clear case of attacking the independence of these two institutions through greater government control—undoubtedly, a retrograde step.
Above all, the decision for the merger has been taken with the approval of the minister and without consulting the NSC. It surely reflects a conscious neglect and lack of basic respect for the independence of the statistical system. One is not questioning the minister’s prerogative as the final decision-making authority, but the process of decision-making without consulting the NSC is questionable. Though the NSC did not exist on the day of the merger, the ministry would have had the full knowledge of the new NSC about to take shape; it was appointed in early July 2019 under the chairmanship of Bimal Kumar Roy.
That the NSC is systematically getting bypassed is evident from the third case of interference in the working of the system. This concerns the refusal by the current government to release the Periodic Labour Force Survey (PLFS) 2017–18 report, even after it was approved by the NSC. The normal practice is for the NSC to approve such reports and the NSSO to release them. For the aforesaid report, there were two technical committees—a PLFS committee for developing the methodology, including the sample design, and a standing committee for PLFS for assessing,inter alia, the annual results. But, the report was not published apparently at the instance of the NITI Aayog, which termed the report as not final but a draft and also “half-baked”(Magazine 2019).
In protest against this, as also against the persistent neglect of the NSC’s role by the government, the then acting chairman of the NSC, P C Mohanan, and a member, J V Meenakshi, resigned on 4 February 2019. In fact, such neglect was voiced by the NSC earlier too. To quote the NSC’s Annual Report 2017–18 (pp 56–57):
The Commission painfully notes that apart from the lack of resources within the Commission, the recommendations of the Commission appear to be of little interest to the Government as no action taken report on these recommendations are even provided to the Commission … even though … concurrence of NSC is mandatory for starting any survey, it is seen that the MOSPI itself has announced surveys/census without any reference to the Commission.
In a series of articles and press interviews after quitting, Mohanan has questioned the NITI Aayog’s tendency for trying to trample upon the autonomy of the statistical system. He has done this with a crusading zeal—all with sound factual reasoning (Mohanan and Kar 2019; Mohanan 2019a, 2019b, 2019c, 2019d).
Structural constraints hindering institutional autonomy: In addition to the concrete instances cited above, there are three unhealthy features inherent in the “restructuring” exercise described in the previous section, which hinder the exercising of autonomy by the statistical system. In the first place, as pointed out above, the decision taken to merge CSO and NSSO was arbitrarily done without involving the NSC. Second, in the hierarchy of the new institutional structure, there is no protection ensured to the CSO and the NSSO for their independence and for safeguarding the integrity of their statistical results.
It is argued that the NSO was also conceived by the Rangarajan Commission. In the first place, that commission had not suggested the merger of the two independent institutions, CSO and NSSO. Far from it; in fact, the commission had made many crucial observations that contradict the proposition of merger. To cite an example, the commission severely criticised the series of steps taken by the then Department of Statistics that had resulted in considerable weakening of the CSO. It argued: “The separate identities of the department of statistics, a purely administrative office, and of the CSO, a professional institution, was thus erased.” This is exactly what the current merger arrangement has done. Thus, the merger is contrary to any of the canons of a well-defined institutional structure as commended by the Rangarajan Commission.
Finally, there is bound to be an inherent conflict of interest between the newly created NSO and the NSC in shouldering their responsibilities. By virtue of its constitution as an independent entity (with a chairman of the status of a minister of state) through a government resolution, the NSC has emerged as a pivotal nodal agency for all core official statistical activities. The creation of the NSC, based on other country experiences, has been for ensuring the independence of the statistical system from government influences in technical results.
The autonomous functioning of the CSO and the NSSO has been ensured by empowering them to produce national-level statistics, after getting their results vetted by independent advisory committees and working groups. It is significant that while the NSC has oversight functions on all field survey reports, it generally keeps an arm’s length distance from the CSO’s national income estimates until these are actually released (Mohanan 2019d).
As for the current design of functional responsibilities between the NSO and the NSC, nothing concrete can be said, for the NSC is just absent in the extant government order for restructuring the system. While the new NSC has been appointed, where it fits into the institutional structure is difficult to discern. That it would be relegated to a “marginal” position is evident from what Srivastava, the CSI-cum-secretary had said in a recent press interview: “NSC is not a statutory body; they can only make recommendations. They have been doing that and we have been taking that into consideration” (Srivastava 2019b). In another context, referring to the need for MOSPI’s control over the CSO and NSSO, Srivastava says: “The National Statistical Commission will continue to work as an overarching body and at an arm’s length” (Srivastava 2019a).1 This ignores the role the NSC now directly plays in approving the NSSO field results.
As for the role of the NSO, we can only speculate, in the absence of any clear details set out after the merger. However, the 2018–19 annual report of the MOSPI provides a synoptic view of what the statistics wing of the ministry is mandated to do. As this wing is restructured into the NSO with the merger of the CSO and the NSSO, the responsibilities expected of NSO are probably what were designed for the statistics wing.
Curiously, these responsibilities specified in the ministry’s annual report consist exactly of those of the NSC, as well as those of the CSO and the NSSO. It is likely that the report has taken the liberty of presenting the functions of the CSO and the NSSO as the functions of the NSO because organisationally, they come under it. But, it is also emphasised that the statistics wing (or now the NSO) “examines the survey reports from the technical angle and evaluates the sampling design including survey feasibility studies in respect of surveys conducted by NSSO and other central ministries and departments.” This directly comes into conflict with the functions formally assigned to the NSC as per the cabinet decision of 10 August 2006, substituting for the governing council of the past. The detailed functions thus expected of the NSC vis-à-vis the NSSO are intended to protect the independence of the NSSO from possible extraneous influences. Mohanan’s resignation from the chairmanship of the NSC has pointedly exposed such possible conflicts of interest between the NSC and the NSO.
The NSC itself, now in operation for over a decade and a half, has no doubt shown some worthwhile results. It has got the Collection of Statistics Act, 2008 enacted, and adopted the Collection of Statistics Rules, 2011. It has also advanced fairly substantially in defining the nature of “core statistics,” drafted a national policy on official statistics, as also on National Statistical Code, and just published a committee report looking into the legislative measures on statistical matters. Along with it, it has prepared a draft bill for establishing the NSC as an independent statutory body with full-time chairperson and members. It is to be a part of “civil society,” independent of and distinct from the government (Menon 2011: 14). Further, the NSC has prepared aReport on the Statistical Audit of All India Index of Industrial Production (IIP) and the Framework for Future Statistical Audits (Sastry 2011: 56–57). It had also completed a major report on unorganised sector statistics (Radhakrishna 2012).
Coordination, Centralisation and Decentralisation
The largest component of Indian statistics is administrative statistics and the administrative division between the central and state governments is based on the constitutional classification of subjects under the union, state and concurrent lists. The tasks of coordination for collection of statistics are thus huge, both at the central and state levels.
Benefits of coordination: In the absence of details on the CSO’s achievements in this coordination task, it is indeed difficult to make any authentic observations. The Rangarajan Commission had expressed displeasure at the way the Department of Statistics was neglecting the CSO’s requirements, which hindered its coordination work (NSC 2001: 451). The situation seems to have undergone changes for the better in recent years, particularly after the setting up of the NSC. Of the two levels of coordination, lateral coordination between the CSO and the central government departments, however, seems to remain weak. The NSC’s 2017–18 annual report has also referred to this: “To some extent, this is due to the failure of the coordination mechanism between statistical advisors in different ministries and the MOSPI”
(p 56). Vidwans (1998: 43) also had criticised thus: “The ministries seem to be deciding things for states independently of the CSO or the states’ DESs (directorates of economics and statistics).”
On the other hand, a review of the available information suggests that the coordination between the CSO and the state DESs is being carried on creditably. The CSO has gradually introduced innovative ideas such as organisation of regular technical meetings and the meetings of Conference of Central and State Statistical Organisations (CCSSOs) for interactions and spread of knowledge as well as coordinated efforts for the production of reliable and timely statistics at the national and state levels. There is also the National Statistical Systems Training Academy (NSSTA) for providing suitable trainings to mid-career level officials.
Thus, the working of the CSO and its coordination efforts have been exhibiting a good example of a centralised agency providing benefits of centralisation for the vertical levels in varied ways: evolving proper statistical standards, extending technical assistance to the statistical cadres in various state-level estimates of gross state domestic product (GSDP), pooling of central and state sample data in which the NSSO staffs are also involved and many other areas of social and economic statistics. The role of CSO’s coordination task has got further expanded for the improvement of the state statistical systems under the World Bank-supported scheme called State Strategic Statistical Plan (triple S), as also the Thirteenth Finance Commission recommendations to improve the state and the district statistical systems. In all of these, the CSO guidelines have played a major role. Alongside, the NSSO has also been actively involved in guiding states on field surveys and pooling of central and state sample studies. It should also be recognised that the presence of the NSC has hitherto played a delicate role in preserving the independence of the CSO and the NSSO without its superior authority being used for any exercise of centralisation or interference in the working of the other statistical institutions.
Centralisation and decentralisation: We would characterise the working of the two institutions in the Indian statistical system, namely, the CSO and the NSSO, as having achieved a delicate balance between centralisation and decentralisation, so far. Such an achievement on part of the CSO, with coordination as an instrument of service, has been explained above. The NSSO, however, as brought out earlier, has a chequered history of its creation. Earlier, its survey work was divided between the NSS in the government and the ISI, Calcutta. The B Sivaraman Committee (1979), which recommended the centralisation of the NSSO in the government, did not consider the alternative of keeping it with the ISI probably because of the inherent perception in the government that the official statistics are a “public good” and that it is the responsibility of the state to produce and disseminate such statistics.
Incidentally, the initial controversy on the subject of centralisation of the sample surveys began with a debate in the early 1950s between D R Gadgil, the author of the famous Poona Schedules of NSS 1950–51 (Rao 2016: 37–38), and P C Mahalanobis. While Mahalanobis recommended the sampling techniques to fill the data gaps for national income estimation purposes, Gadgil on the other hand was doubtful about “whether information can really be obtained in India by random survey,” given the absence of familiarity of those employed for the survey with the rural conditions, among other things. He thought that the gaps relating to national income estimation “should be filled by a series of related but separate bits entrusted to ministries, state bureaus and competent non-official agencies” under the guidance of the national income committee, but “not be attempted to be done in one uniform way by one central agency.2 With his experience of large-scale surveys, Mahalanobis placed greater emphasis on nation-wide sample surveys. Finally, Mahalanobis won the debate and thus the NSSO was born in 1970 as a centralised, government-run agency.
The NSSO inherited a vast array of statistical application-oriented research done at the ISI. Therefore, there were some misgivings regarding the possible inadequacy of research support for sampling methodology in the NSSO when separated from ISI. To quote one such doubt:
Indeed, the separation of the NSSO from the ISI has the potential danger that in the event of less than adequate methodological research support, the mode of operability as well as capability might have to be compromised. Of course, the NSSO may have, by now, an adequate number of research workers to guard against such probable erosion. (Sen 1998: 79)
But, indeed the statistical system has established some alternative protection against such eventualities, such as the system of a governing council manned with experts initially, the subsequent appointment of an expert NSC, and the institution of study groups to examine the working of the entire NSSO and its methodological designs. There have also been national-level seminars seeking comments from academicians on the survey results. One of the committees suggested improving the standard of theSarvekshana, the NSSO’s journal (see Rao [2012] for details). It is noteworthy that all of these suggestions have been implemented over time directly or indirectly. Just to cite a studied observation of A Vaidyanathan in this respect:
The National Sample Survey, set up as an independent professional organisation to generate data on aspects not covered by government agencies, has made valuable contributions in improving data. It has demonstrated the effectiveness of sample survey techniques as a cost-effective and reliable way of getting information. In some areas—consumption, employment and assets—it is the only source of detailed and comparable data over time. The fact that its programmes, survey design and field producers are reviewed and decided in close interaction with users and academics, is its strong point. (Vaidyanathan 1998: 35)
Thus, so far, the CSO as well as the NSSO has treaded a balanced path insofar as issues of centralisation and decentralisation, intermingled with coordination, are concerned. However, a few recent developments have raised serious concerns about crude centralisation of national statistics management and dissemination. This concern arises from the merger of the CSO and NSSO, the two hitherto reasonably independent institutions, as explained earlier. Instead, the earlier arrangement of the statistics wing continuing with its administrative control over different wings in the ministry and not disturbing the independence of the institutions carrying on with their technical work, would have been very apt.
Unintended result of centralisation: One example wherein the enthusiasm shown by the CSO to revise the national income series from 2004–05 to 2011–12 by applying new methodology and using new data sources as per the 2008 United Nations System of National Accounts (UNSNA), has resulted in undermining the role of the state statistical bureaux in the estimation of GSDP as per revised series.
For the 2011–12 series, the CSO has made a quantum leap in adopting new data sources (MCA21 corporate sector data, improved coverage of local bodies, incorporation of new NSSO surveys for the unincorporated enterprises) and new methodology (effective labour input method)—all of which were technically sound and the logic of their introduction indisputable (Shetty and Rajakumar 2017). But as a result, the dependence of states for their GSDP compilation has shot up radically as the CSO’s national-level estimates are apportioned amongst states by some suitable indicators. Such induced centralisation has been severely criticised on many grounds, mainly because such centralised method of GSDP estimation producing a very damaging effect for the DES staff as their concerned division would not have much work and “once their systems, processes and data sources are dysfunctional and dismantled, it would take a long time to reconstruct them and build the necessary competence” (Dholakia and Pandya 2017: 136).
While all that has been said above is correct, the resulting drawback cannot be attributed to a conscious process of centralisation of data compilation at the CSO level. The central agencies with better knowledge and skill levels cannot ignore their responsibilities to produce estimates based on available information and as per the prescriptions of UNSNA. At the states level, there is serious administrative laxity and chances of their falling in line in the foreseeable future are remote.
Be that as it may, instead of only blaming the states, the administrative apparatus concerned with the statistical system at the centre has also to rise to the occasion. It has to look inward and ask itself whether it has made an attempt to install a permanent statutory machinery like the NSC with adequate powers and resources as proposed by the Rangarajan Commission, so that it can protect the independence of the statistical system. How many of the valuable recommendations of that commission have been implemented3 and above all, how much of the proposed freedom and independence have been allowed to the key institutions in the statistical system? Answers to these questions will help us resolve the issues being debated and also help us achieve a healthy compromise between centralisation and decentralisation.
Private Agencies
Considering the series of constraints faced by the public administration or even otherwise, a question arises whether there is scope for involving private and other non-governmental agencies in producing official statistics. The subject was debated several times over in the past and the consensus was that official statistics is a “public good.” Also, an overwhelming part of the official data sources is administrative statistics, which only an official agency can procure for further consolidation.
As for the field survey data, the subject had received fresh attention when the World Bank under its 1998 project on “Modernisation of the Statistical System in India” proposed a massive expansion of surveys through the NSSO to replace the failing administrative statistical system. This proposal was severely criticised by S M Vidwans, who was a member of the Rangarajan Commission, and by the commission itself. It was argued quite admirably that a massive survey-based system cannot replace the nuts and bolts type of input requirements for macro statistics that have to come from the day-to-day administrative machinery at the central and state levels (Vidwans 2002). The Rangarajan Commission further argued that such dependence on massive NSSO surveys will only “divert attention from the solution of the real systemic problem of the decentralised Indian statistical system … and result in quick processing of data, deficient in quality, coverage, timeliness, accuracy or precision” (p 451 of Volume II).
The Rangarajan Commission had also questioned the associated proposal to engage straightway the non-governmental organisations (NGOs) and other reputed private agencies as official data collection agencies on the ground “that the government cannot abjure its primary duty to collect data on as many aspects of people’s life as possible, that there is no guarantee that data collection by private agencies would not be subject to their own biases” (NSC 2001: 448).
That does not mean that in the annals of data gathering, the agencies outside the government machinery do not have any role. Such agencies can play useful role in many respects. We have in India the examples of the Centre for Monitoring Indian Economy (CMIE) undertaking the compilation of corporate finances data and field data on employment and unemployment, etc, and the National Council of Applied Economic Research (NCAER) undertaking field surveys on household income, consumption and saving, which have all made valuable contributions in generating large-size statistical works supplementing official data.
Even so, the question before us is whether such data can be a substitute for official statistics. Our considered view is that it cannot be, and as the Rangarajan Commission has opined, the government cannot abjure its responsibility of producing official statistics as a “public good.”
Concluding Assessments
The most noteworthy aspect of the development of statistical institutions in this country has been their sense of independence and integrity. This philosophy has been steadfastly guarded until recently. The various institutional arrangements made such as the advisory committees for the CSO and/or the governing council until the setting up of the NSC in 2006, and the working groups and standing committees, along with scrutiny by the NSC itself for all rounds of the NSSO thereafter, have been a commendable attempt at ensuring the independence and integrity of the system.
But, recording such an enviable history of autonomy and integrity, one is disheartened to report that certain recent developments suggest that the governmental machinery appears to have attempted to compromise with the independence and integrity of the statistical institutions for political gain. More importantly, the motive behind the merger of the CSO and NSSO into the NSO, directly under the control of the ministry raises serious suspicion. The alleged absence of control of earlier governments over these institutions and their independent functioning seem to have posed a challenge for the current government’s desire to control these and therefore, they have been placed directly under the MOSPI’s grip. The most logical step should have been the earlier arrangement of the statistics wing continuing with its administrative control over different wings in the ministry and not disturbing the independence of the institutions carrying on with their technical and professional work.
It must be emphasised that the contention here is not to establish that the ministry will interfere with all statistical exercises, but to remind that the merger of the two institutions under a unified wing placed directly under the government control gives the freedom to the government to interfere with the statistical results as and when exigencies arise for its image. It can even do so without stirring any public attention. As the recent experiences clearly show, such interferences under the present dispensation of independence and autonomy cannot but be exposed to public glare.
The recent developments do not question the competence and the integrity of the official statistical cadre in India. Their reputation has proved that they are unquestionable. What these developments have done though is just to facilitate government interference in their statistical activities whenever the government wants to do so, which does of course tarnish their otherwise sound reputation of competence and impartiality.
Notes
1 Here we would like to draw attention to the government’s press release dated 19 December 2019, seeking comments/suggestions on a proposed draft of the NSC Bill 2019. The bill envisages a statutory NSC with a full-time chairperson and five whole-time members, instead of the current system of part-time chairpersonship and membership. We assert that though the statutory NSC is a welcome step, but the draft bill does not obviate to any extent the blatant way the professional autonomy and independence of the statistical institutions like the CSO and the NSSO have been compromised by merging the two institutions into the NSO, which remains directly under the ministry.
2 See Rao (2012, 2016) for further details.
3 There is no systematic information in the public domain about the implementation of the Rangarajan Commission report. Only the following some bland information is available: The Commission’s Annual Report for 2016–17 writes: “The Commission chaired by C Rangarajan in its comprehensive report (August 2001) made 623 recommendations for the improvement of the Indian statistical system. During a review of its 60th meeting held on September 2013, out of 623 recommendations in all, 147 recommendations have been taken as implemented, 9 have been rejected/dropped and 467 are still pending.” On a subsequent closer scrutiny, it has been revealed that 116 are still pending. The review hardly provides any insight into which of the substantive recommendations have been implemented.
References
Dholakia, Ravindra H and Manish B Pandya (2017): “Critique of Recent Revisions with Base Year Change for Estimation of State Income in India,”Journal of Indian School of Political Economy, Vol XXIX, Nos 1 and 2, January–June.
Iyer, P V (2018): “Three Years Ago, Key Statistics Panel RevisedUPA Growth Up,NITI Aayog Rejected It,” Indian Express, 7 December.
Magazine, A (2019): “NITI Aayog Calls Adverse Job Report a Draft, Ex-stat Head Says No Its Final,” Indian Express, 1 February.
Menon, Madhava N R (2011):Report of the Committee on Legislative Measures in Statistical Matters as Adopted by the National Statistical Commission, National Statistical Commission Secretariat, MOSPI, 27 October.
MOSPI (2019): Restructuring of Indian Official Statistics System, Order No: M-12011/1/2019-CAP, Ministry of Statistics and Programme Implementation, 23 May.
— (2019):Annual Report 2018–19: Contribution of Different Sectors to Gross Value Added in 2017–18, Government of India.
Mohanan, P C (2019a): “The Jobs Report Cannot Be Called a Draft, It Is Final Once I Approve It,” & “You Can’t Involve an Outside Agency Like NITI; It’s a Political Body,”Mint, 11 February.
— (2019b): “Because Data Is a Public Good,”Indian Express, 12 February.
— (2019c): “The Pillar Stands,”Indian Express, 2 March.
— (2019d): “We Felt NSC Is Not Able to Contribute to Objectives for Which It was Set Up,”Mint, 31 January.
Mohanan, P C and Aloke Kar (2019): “Data in a Narrative-driven Debate” & “Autonomy of Statistical Agencies,”Business Standard, 13 and 14 February.
NSC (2001):Report of the National Statistical Commission (Chairman: Dr C Rangarajan), Volume II, MOSPI, National Statistical Commission, August.
NSO (2019):Annual ReportofPeriodic Labour Force Survey (PLFS): 2017–18, National Statistical Office, Government of India, New Delhi, May.
Radhakrishna, R (2012):Report of the Committee on Unorganised Sector Statistics (Chairman: Prof R Radhakrishna), National Statistical Commission, January.
Rao, T J (2012): “Origin and Activities of NSS, Government of India,” Conference Paper, Invited paper in the Proc ofInternational Conference on Large Scale National Surveys,
Rajshahi University, Bangladesh, 18–19 October, [https://www.researchgate.net/publication/319292092_ORIGIN_AND_ACTIVITIES_OFNSSO_GOVERNMENT_OF_INDIA/].
— (2016): “Mahalanobis: Chief Architect of National Sample Survey and Related Contributions,”Sarvekshana—100th Issue, Journal of National Sample Survey Office, MOSPI.
Rajakumar, J Dennis (2018): “Question Mark over CSO’s Independence,”Hindu Businessline, 10 December.
Sastry, N S (2011): Report on Statistical Audit of All India Index of Industrial Production and Framework for Future Statistical Audits, National Statistical Commission Secretariat, MOSPI,
31 May.
Sen, Pranab Kumar (1998): “Fifty Years of the NSSO—An Appraisal,” Indian Statistical System Golden Jubilee of Indian Independence, Commemorative Volume 1, Department of Statistics, MOSPI, June.
Shetty, S L and J D Rajakumar (2017): “New National Accounts Series: A Review and Highlights of Crucial Issues for Debate,”Journal of Indian School of Political Economy, Vol XXIX, January–June, Nos 1 and 2.
Srivastava, Pravin (2019a): “In a Major Statistical System Overhaul, Government Plans to Merge CSO, NSSO,”Business Standard, 26 May.
— (2019b): “Why Government has Revamped the Country’s Statistical System,”Mint, 5 June.
Vaidyanathan, A (1998): “India’s Statistical System 50 Years After Independence,” Indian Statistical System Golden Jubilee of Indian Independence, Commemorative Volume 1, Department of Statistics, MOSPI, June.
Vidwans, S M (1998): “Whither Statistics?” Indian Statistical System Golden Jubilee of Indian Independence, Commemorative Volume 1, Department of Statistics, MOSPI, June.
— (2002): “Indian Statistical System at the Crossroads I, II & III,”Economic & Political Weekly, 14, 21 and 28 September.
