Prime Minister of India - Dr. Manmohan Singh

PM's speech at Platinum Jubilee of Indian Statistical Institute

"I am truly delighted to be here today to participate in the platinum jubilee celebrations of this great institution. As a student of economics, I know how much the so-called 'dismal science' owes to statistics, especially when it comes to proving that things are in fact not so dismal! There is hardly any academic discipline today that is not in some way indebted to statistics, whether it is medicine, engineering, finance, sociology, anthropology or information technology.

Seventy-five years ago a lamp was lit by Professor Prasanto Chandra Mahalanobis, the doyen of Indian Statistics, and that lamp continues to burn even today. A distinguished lineage of Indian statisticians who have studied and taught here, including Professors J.N. Sengupta, H.C. Sinha, R.C. Bose and C.R. Rao, to mention just a few, have greatly enriched the discipline of statistics. I pay tribute to their memory.

The Indian Statistical Institute stands apart as a unique institution devoted to the research, teaching and application of statistics to natural and social sciences. It was in recognition of this role that it was declared an `Institution of National Importance' by an Act of Parliament in 1959. Like so many institutions of national importance in our country, the ISI also blossomed and grew during the Nehruvian era. We owe it to the vision and leadership of Panditji that he helped build the foundation of intellectual effort in modern India. He offered space and opportunity to a galaxy of great researchers and teachers who contributed to this process.

Professor Mahalanobis, widely acclaimed as the father of Indian statistics, was a far-sighted thinker. He realised the importance of statistics and statistical analysis and research in effective policy formulation. Therefore, even as the institute blossomed in the areas of theoretical research, Mahalanobis recognised the importance of training to equip statisticians with application skills.

The first exploratory sample survey in India for improving the estimates of jute crop was undertaken under Prof. Mahalanobis' technical guidance as early as in 1937. This study won the praise of Ronald Fisher who, I believe, wrote to the then Government stressing the importance of such studies, saying, "...everything depends on the future of the Statistical Institute."

The Institute went on to do a number of other sample surveys relating to economic and social conditions, especially in Bengal. The pioneering work on crop yield estimation done by the Institute attracted the attention of Pandit Jawaharlal Nehru. He sought comprehensive information relating to social, economic and demographic characteristics and wanted sample surveys conducted on a countrywide basis.

Thus was born in 1950, the first National Sample Survey (NSS) that is now a massive operation. Today, the NSS covers nearly 2.5 lakh households and is an outstanding example of the effectiveness of the sample survey technique as a cost effective and reliable way of getting information. In areas such as consumption, employment and assets it is the major source of detailed and comparable data over time. Pandit Nehru's invitation to Professor Mahalanobis to work as Honorary Statistical Adviser to the Union Cabinet, brought ISI closer to national planning and policymaking activities.

The 1950s and 60s were a unique period in our developmental history. There was great interaction between officials in government and scholars in universities. There was a two-way flow of talent between institutions like the Delhi School of Economics and the Indian Statistical Institute and the various ministries of our government. This interaction enriched the quality of academic research, making it more policy-oriented, and also contributed, I believe, to creative thinking within government.

The Indian statistical system, built from scratch by stalwarts, many of who were also closely associated with ISI, is today the envy of many developing countries. We are one of the few developing countries to conform to the discipline of the IMF's Special Data Dissemination Standards under which India is committed to providing quality data in a host of areas relating to the real, financial and external sectors of the economy and at regular intervals.

The setting up of the National Income Committee in 1949, NSS in 1950, CSO in 1951 and the NSSO in 1970 are important institutional landmarks in the evolution of the Indian statistical system. We have an elaborate infrastructure to capture a wide variety of data for an economy that is both large and decentralised.

However due to over�dependence on the administrative set up and traditional records, the system has not been able to keep pace with changing times. Liberalisation has also ushered in significant structural changes in the economy - changes that need to be captured by the statistical system. Unfortunately, despite impressive and commendable progress since Independence there are growing concerns about the reliability, timeliness and accuracy of our statistical data. Frequent revisions of important data series have added to the erosion of credibility.

Economic theory can at best establish the relationship among identified variables. But for policymaking, we need to be able to quantify this relationship. This, in turn, calls for reliable data and on a timely basis. Statistical inference is an important part of any application. However inference will be useless if the basic data are faulty or inaccurate or unreliable.

I must add, however, that this concern with the quality of data is not unique to India. Nor is it new. In fact there is a reference to a judge in British India admonishing an over �zealous civil servant who quoted official statistics to him with the words, "Governments are very keen on amassing statistics - they collect them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But what you must never forget is that every one of these figures comes in the first instance from the village chowkidar who just puts down what he damn well pleases!"

In recent years many countries including U.K., Australia and Canada have focused attention on improving their statistical systems. `Building Trust in Statistics' was the underlying theme of their efforts to revamp the statistical system. In India, apart from questions about the quality of data, there are problems such as data gaps, duplication leading to conflicting statistics and inordinate delays in transmission and publication of data.

In agriculture for instance, large gaps exist with reference to the output of fruits and vegetables and other minor crops as also estimates of meat products and fish. This is in addition to the perennial problem of deterioration in the quality of data on crops. As a policy maker, I am faced with conflicting data on crop estimates of important products such as wheat and sugarcane.

As far as industry is concerned, the Index of Industrial Production fails to fully capture our vibrant small-scale and informal sector. Services now account for 60% of GDP. Yet services too are not fully captured in our national income statistics. The task becomes difficult given the vastness of the sector, its heterogeneous nature as well as fast-changing composition with the frequent emergence of new services and the exit of obsolete ones.

Another key issue has been the discrepancy in the estimates of the same variable from two different sources. The classic example is the difference between the National Accounts and National Sample Survey estimates on consumption expenditure. The present system has also not been able to provide adequate information on basic socio-economic indicators that are crucial for micro level planning. No reliable data is available on many aspects of healthcare and education, two crucial areas for our economy.

Similarly, the deregulation of financial markets has accelerated the pace of financial innovation and brought home the need for regular and timely flow of quality financial statistics. In the context of recent financial crises, traditional issues in financial data such as timeliness in dissemination, accuracy, transparency, harmonization, international comparability, etc. have come into sharper focus.

Changes in the prices of goods and services affect different segments of the population differently. Thus measuring prices and their rates of change over time has become crucial to almost every economic issue from the conduct of monetary policy to measurement of poverty. Thus data on prices need to be beyond all doubt and must be in line with changing consumption patterns. The services sector, for instance, contributes significantly to the GDP, and the need for the computation of a separate price index for services has become urgent.

Overcoming these deficiencies will call for reforms in the administration of the Indian Statistical System and upgrading its infrastructure. The National Statistics Commission, under the chairmanship of Dr C Rangarajan, which went into these issues pointed to the imperative need to improve credibility, timeliness and adequacy of our data system. It made a number of recommendations including reform of the structure of the Indian Statistical System and upgrading its infrastructure to ensure autonomy, improvement in the collection of datas, exploration of alternative techniques for collection of datas, identification of new data series that may need to be generated in keeping with the changing economy, and, evolving appropriate methodologies for collecting new data in relation to new data needs.

Our government firmly believes in transparency and openness. Reliable and timely statistics is a basic requirement for this. Accordingly, we set up a nodal body - the National Statistical Commission - to oversee all core official statistical activities. The new body - the National Statistical Commission, headed by the eminent economist Prof Suresh Tendulkar, has assumed charge in July this year

The Commission has a wide mandate. It would oversee the design and implementation of strategies for collection, analysis and dissemination of core statistics as per a standardized advance release calendar. The Commission would also oversee a comprehensive nationwide programme for capacity creation in collection and analysis of statistics. The Commission's mandate will also include improving the quality of statistics, enforcing appropriate standards, encouraging the use of statistical analysis in policy formulation and increasing public awareness of official statistics.

It is indeed a matter of pride that the ISI has continued to deepen and extend its relationships with the social, natural and applied sciences. It has provided a wonderful platform for a meaningful interaction between the government, industry, top researchers, high quality teachers and motivated students. The canvas is vast. I can but draw attention to the quality of faculty and visitors. For an organization to have hosted the likes of JBS Haldane and George Akerlof, is indeed remarkable. What makes it even more remarkable is the ISI's ability to attract relatively younger faculty away from lucrative assignments abroad. As an economist, I am impressed by ISI's close association with the best of contemporary Indian economists, like Kaushik Basu, T.N. Srinivasan and B.S. Minhas. I am told that the mathematics and statistics units are equally distinguished. It is a matter of pride for all of us.

In recent years this Institute's involvement in the planning process and for strengthening the statistical system of the country has been on the wane. The Institute needs to participate and play a major role in upgrading the quality of data generated by various Ministries of the Government of India and by the State governments. It must focus on research in the fields relevant for policy formulation and more specifically, in the domain of official statistics. The ISI should help in the technical upgradation of data generated by the various Ministries of the Government of India and the State Governments. It must improve sample survey methodology by updating and renovating techniques of data collection and imparting training in statistical methods to key personnel in their organizations.

The tasks of perspective planning and survey designing assigned to the institute in the past have not been replaced by any similar assigned task on a regular basis. A formal arrangement permitting the institute's collaboration with government agencies involved in collection, compilation and analysis of official statistics might help in the improvement of the country's statistical system. The Institute could consider setting up an Official Statistics unit that would coordinate between the Central Statistical Organisation and researchers at ISI in matters of research. The Institute could consider supporting collaborative research between ISI and official statisticians on topics relating to 'official statistics'.

As in the case of other premier teaching & research institutes like the IIT's & IIM's, this Institute needs to diversify to other parts of the country to cater to the growing needs of our country. This is the opportune time for the Institute to think along these lines.

A judicious balance has to be envisaged between individual academic work on truly fundamental problems and work that calls for a greater engagement with the social and economic problems facing the country.

I look forward to the ISI continuing to contribute to the improvement of our official statistical system and to the development of the discipline. ISI must regain its reputation as an institution of world-class excellence. In the world of today, competition in any field is competition in knowledge and that is why we need institutions of excellence like ISI. I am sure that the lamp lit by Professor Mahalanobis 75 years ago will continue to burn bright in the future.

I wish you all and the Indian Statistical Institute the very best in the years to come. I am sure that the best is yet to come."