MoSPI Proposes Seasonal Adjustment Methodology To Enhance Industrial Data Reliability
SDG 8: Decent Work and Economic Growth | SDG 9: Industry, Innovation and Infrastructure | SDG 17: Partnerships for the Goals
Ministry of Statistics & Programme Implementation (MoSPI) | National Statistical Office (NSO)
The Ministry of Statistics and Programme Implementation (MoSPI) has proposed the official introduction of seasonally adjusted Index of Industrial Production (IIP) series. This move aligns India with international recommendations from the UNSD, OECD, and Eurostat, which advocate for deseasonalisation to enhance the interpretability of high-frequency data. While these agencies recognize seasonal adjustment as essential for identifying underlying economic signals, the paper acknowledges inherent trade-offs, including subjectivity in model choice, the risk of misleading signals if poorly executed, and the significant human and computer resource burden required to maintain dedicated adjustment units.
Addressing Unique Indian Seasonality Factors Industrial production is influenced by a composite of climate and institutional events that repeat regularly. MoSPI identifies several core drivers of seasonality:
Climate and Calendar: Weather patterns and the varying number of working days in a month can distort short-term comparisons.
Moving Holiday Challenges: A significant hurdle in the Indian context is the moving holiday effect. Unlike fixed holidays, major festivals such as Diwali, Holi, and Dussehra follow lunar cycles and shift across Gregorian months. These shifts affect social behavior and cause a loss of working days, requiring specialized calendar regressors to avoid biased trend estimates.
Methodological Framework and Quality Control The proposed process involves decomposing a time series into three unobservable components: the trend cycle (long-term direction), the seasonal component (systematic movements), and the irregular component (unpredictable noise).
Decomposition Models: MoSPI will utilize additive models (where components are independent) or multiplicative models (where seasonal variations change with the level of the series).
Quality Benchmarks: Effective adjustment requires the total removal of residual seasonality so that no estimable seasonal effects remain in the adjusted series.
Outlier Treatment: To prevent “spurious seasonality,” the framework includes the identification and correction of outliers, such as one-time shocks (Additive Outliers) or permanent changes (Level Shifts), particularly those observed during the COVID-19 period.
What is “Deseasonalisation” and why is it distinct from “Trend” analysis? Deseasonalisation (or seasonal adjustment) is the process of removing only the predictable, repeating annual patterns from a dataset. The resulting seasonally adjusted series still contains both the long-term “Trend” and the unpredictable “Irregular” noise (such as strikes or sudden shocks). While deseasonalisation makes short-term movements clearer, it is the Trend-Cycle component that provides the smoothest view of the economy’s actual direction.
Policy Relevance
The introduction of seasonally adjusted IIP represents a transition toward more sophisticated, “news-oriented” economic monitoring.
Granular Economic Interpretation: Removing seasonal “noise” allows policymakers to detect business cycle turning points much earlier, facilitating timely interventions.
Evidence-Based Governance: By adjusting for moving holidays like Diwali, the government can distinguish between a genuine factory slowdown and a temporary dip in operational days.
Global Credibility: Adopting these rigorous methodologies enhances the comparability of India’s industrial data with major global economies, supporting a more transparent investment climate.
Follow the full paper here: Discussion Paper 3.0: Seasonal Adjustment of Index of Industrial Production