The Reserve Bank of India organised the 7th Statistics Day Conference on August 30, 2013. The theme of the conference was 'Challenges to Policy Making and Evolving Role of Statistical Analysis'. Statistics Day is observed in India to commemorate the sterling contribution of Prof. Prasanta Chandra Mahalanobis to statistics and the Indian statistical system.

Distinguished statisticians and academicians from India and abroad, such as, Prof. Bimal Roy, Director, Indian Statistical Institute, Kolkata; Prof. Richard Smith, University of Cambridge, UK; Prof. B. L. S. Prakasa Rao, C. R. Rao Institute of Mathematics, Statistics and Computer Science, Hyderabad; and Prof. Subhashis Ghosal, North Carolina State University, spoke in the conference. The conference deliberated upon the role of statistical measurement and analysis for dealing with policy challenges, use of survey data and recent advances in high-dimensional data analysis. Recent initiatives by the Reserve Bank in development of indicators for macroeconomic measurement were also presented.

Dr. D. Subbarao, Governor, in his inaugural address, alluded to a host of conceptual issues in statistical analysis in the course of policy making and dilemmas faced in resolving them during the last five years. In his remarks, he raised the following five conceptual issues: (i) Inflation measure for monetary policy in India: WPI or CPI? (ii) Potential output, threshold inflation and Taylor Rule for monetary policy reaction function; (iii) Neutral interest rate and assessment of monetary policy stance; (iv) Equilibrium exchange rate - Why no single statistical measure is acceptable as a consensus reference value?, and (v) Financial stability and stress indicators - way forward. Through these issues, he highlighted the formidable challenges confronting statisticians in a world which is becoming increasingly integrated and complex. Simultaneously, the demands on statisticians are increasing as policy makers are looking towards them for robust and reliable technical analysis to guide policy making

Prof. Bimal Roy, in his keynote address, reminisced the role and contributions of the Indian Statistical Institute in the development of Statistics and Statistical Systems in India. He recounted in particular the fundamental contributions of internationally acclaimed statisticians of ISI in the fields of sample survey, design of experiments, theory of estimation, inference, probability and stochastic processes. He highlighted the role of statistics in implementation of data security including algorithm for data encryption. He emphasised that internal testing and monitoring of these aspects would be critical for the economy, especially for the financial sector.

Prof. Richard Smith presented an econometric framework of using qualitative survey data for forecasting of economic activity. Categorical responses of the firms were linked with official data through ordered discrete-choice models and a statistically efficient composite indicator was developed. Empirical results based on the proposed indicator showed its relative advantage of producing early estimates of output than traditional indicators.

Prof. Prakasa Rao highlighted the recent advances in statistical analysis of high-dimensional data, which is characterized by large number of parameters as compared to the number of observations. As the classical statistical methods are not applicable to analyse such data, alternative methods were suggested to model the intrinsic heterogeneity of big and high-dimensional data.

In spite of the very high complexity of large-dimensional data, a key feature that allows valid statistical analysis is its property of sparsity. How to use this characteristics in a Bayesian framework was the essence of Prof. Ghosal’s talk. From the computational perspectives, he proposed a new technique based on Laplace approximation for structure detection and demonstrated its usefulness for the analysis of high-dimensional financial data.

Dr. Urjit R. Patel, Deputy Governor highlighted three major areas of research as (i) development of models for forecasting of macroeconomic variables and policy simulation which captures the dynamics and the stochastic behaviour of the economic agents; (ii) explore the scope to integrate the results of surveys on industrial outlook, inflation expectations, consumer confidence and credit conditions in models to enhance their utility for policy, and (iii) strengthen our research capabilities for analysing financial market data, where the linkage with real sector constitutes an important feedback channel.

Earlier, while welcoming the participants, Shri Deepak Mohanty, Executive Director, noted that in the Reserve Bank, we treat statistics as a public good. It is necessary not only to aid our decision making but also for empowerment of citizens. He emphasised the need for harmonisation of data definition across various returns submitted to RBI and adoption of straight through processing by reporting entities without manual intervention for submission of information to maintain data integrity. He also underscored the need for enhancing the scope and coverage of corporate finance statistics and further development of asset price statistics.

The deliberations in the conference provide input for crafting the agenda of future statistical work in the Reserve Bank.

Alpana Killawala

Principal Press Relations Officer

Statistics as a public good: harmonised data and advanced analysis to strengthen monetary and financial policy decision making. Statistical measurement and analysis are central to policy formulation, with key issues including choice of inflation aggregate, potential output and threshold inflation for policy rules, neutral interest rate assessment, equilibrium exchange rate benchmarking, and development of financial stability and stress indicators. Methodological responses emphasised integrating qualitative survey data into forecasting, adopting alternative methods for high dimensional data using sparsity and Bayesian structure detection, and enhancing macroeconomic and financial market modelling. Institutional recommendations called for treating statistics as a public good, harmonising data definitions, straight through processing for reporting, and expanding corporate finance and asset price statistics.