Business Cycle Measurement in India

Abstract

This chapter presents the business cycle chronology for the Indian economy. Two distinct phases are analysed: the pre-1991 period when the cycles were mainly driven by monsoon shocks and the post-1991 phase where we see the emergence of conventional business cycles driven by investment-inventory fluctuations. The chapter sheds light on the economic conditions that shaped the nature of cycles in the two phases. The concluding section of the chapter presents an overview of the economic conditions post-2012.

Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Appendices

Appendix 1: Detection of Turning Points Using the Dating Algorithm

The Bry-Boschan (BB) and Harding Pagan (H-P) algorithms find the turning points as follows:

• The data is smoothed after outlier adjustment by constructing short-term moving averages.

• The preliminary set of turning points is selected for the smoothed series subject to the criterion described later.

• In the next stage, turning points in the raw series are identified taking results from smoothed series as the reference.

The identification of turning point dates is done subject to the following rules:

• The first rule states that the peaks and troughs must alternate.

• The second step involves the identification of local minima (troughs) and local maxima (peaks) in a single time series or in y_t after a log transformation.

• Peaks are found where y_s is larger than k values of y_t in both directions.

• Troughs are identified where y_s is smaller than k values of y_t in both directions.

• Bry and Boschan (1971) suggested the value of k as five for monthly frequency which Harding and Pagan (2002) transformed to two for quarterly series.

• Censoring rules are put in place for minimum duration of phase (from peak to trough or trough to peak) and for a complete cycle (from peak to peak or from trough to trough).

• Harding and Pagan identify minimum duration of a phase to be two quarters and the minimum duration of a complete cycle to be five quarters.

• For monthly data, the minimum duration is 5 months and 15 months for phase and cycle, respectively.

• The identification of turning points is avoided at extreme points.

Appendix 2: Recent Changes in the Indian GDP Measurement

The business cycle chronology presented in the preceding section is based on the GDP series with base year 2004–2005. In 2015, the Indian Central Statistical Office (CSO) introduced the new series of National Accounts Statistics with the base year 2011–2012, replacing the earlier series with 2004–2005 as the base year. In contrast to the earlier episodes of base year changes, this update was marked by changes to the methodology and data sources. The methodological changes were implemented to align the Indian National Accounts Statistics with international standards recommended by the System of National Accounts (SNA) 2008. The state of the economy is measured using gross value added (GVA) at basic prices, in place of the earlier practise of measuring it using GDP at factor cost.

The key methodological refinement is seen in the manufacturing sector where the gross value addition is computed using comprehensive data sources such as the MCA-21.²¹ Despite methodological improvements, the revised series has attracted considerable debate amongst academicians, policy-makers and other stakeholders (Sapre and Sinha 2016; EPW 2015; Nagaraj 2015a, b). In this section we discuss some concerns with the new series.

Changes in the Sub-sectors’ Growth

Table 11 shows that there are striking changes in the subsectors’ growth rate for the two intermittent periods when we have data from both the series. For instance, the growth rate of the gross domestic product (GDP) for 2013–2014 according to the new series was 6.6%, compared to 4.7% in the earlier series. The greatest discrepancy is seen in the growth rates of manufacturing sector. According to the new series with base year 2011–2012, the growth rate of manufacturing was 5.3% in 2013–2014, while the old series shows a contraction in the manufacturing sector for the same year.

Table 11 GDP and subsectors’ growth rate

Disconnect Between the High Frequency Indicators and the Sectoral GVAs

Due to changes in the methodology, the high-frequency indicators which conventionally mapped the trends in GDP subsectors no longer seem to be in sync with the new subsectors’ GVA. Figure 12 shows the discordant trends in the high-frequency indicators and the related subsectors of GVA. Figure 12 shows that IIP is at odds with the movement of GVA in the manufacturing sector. Similarly bank credit data does not seem to be in sync with the new GVA of finance, insurance and real estate.

Fig. 12

Comparison with high-frequency indicators

Choice of Deflator

The estimates of real GVA in most advanced economies are arrived at using double deflation. In this method, nominal outputs are deflated using an output deflator, while inputs are deflated using a separate input deflator. Then, the real inputs are subtracted from real outputs to derive real GVA. But in India things are done differently. Here, we compute the nominal GVA and then deflate this number using a single deflator.

If input and output prices are synchronous, both approaches will give similar results. But if the two price series diverge—as they have for the past few years in India—single deflation can overstate growth by a big margin (Sengupta 2015).

Issues with Manufacturing Gross Value-Added

The manufacturing sector has been at the centre stage of the GDP debate.

• 1. Enterprise vs establishment approach: In a major change in methodology, the data collection for GVA computation shifted from establishment or factories to enterprise or firms. Conceptualising value addition at the enterprises level without clarity on measures of costs and output could lead to misleading estimates of GVA (Sapre and Sinha 2016). The activities of firms can be much more diverse than those of factories, and if all these go into the calculation of GVA, it could inflate the estimate of output.

• 2. Blowing up of GVA: GVA calculation involves identifying a set of “active companies” that have filed their annual returns at least once in past 3 years. The problem is that for any given year, information from several active companies remains unavailable till a cut-off date of data extraction. In such a case, the GVA of available companies needs to be “blown-up” to account for the unavailable companies. Literature has commented on a number of issues with the blowing-up method.

  The year-wise number of available and active set of companies in the manufacturing sector is not publicly available, so the extent of blowing up is not known. Some experts have criticised the methodology of blowing-up. The critical input is the “blowing-up factor” which is the inverse of the ratio between the paid-up capital (PUC) for the available companies and that for the active set as a whole. Nagaraj (2015a, b) argues that this is inappropriate since a number of the companies in the “active set” could be shell companies existing only on paper. This could overestimate gross value added of the manufacturing sector.

• 3. Discrepancies in the underlying data sources: For the manufacturing sector, the GVA is derived from a combination of MCA-21 numbers, index of industrial production (IIP) estimates and estimates of the unorganised sector from the Annual Survey of Industries (ASI). While the MCA-21 is a new database, the base year for the IIP data is still 2004–2005. Also the data obtained from MCA-21 follows an “enterprise” approach as mentioned earlier, but the data obtained from ASI follows the old “establishment” approach. This could lead to misleading estimates of the GVA numbers (Sengupta 2015).