New economic data and their consequences
The release of new data on Index of Industrial Production (IIP) and wholesale price index (WPI) is a hugely positive development. Of course, we have not reached a resting point. Upgrading data is everywhere an Alice-in-Wonderland-ish business of catching up with fast-changing reality, and the new indices are based on the Indian economy as it was six years ago, in 2011-12. More dynamic updating is therefore essential, an effort that is already and commendably under way.
Much has been written about how the new data have been constructed. Here, I want to focus on their consequences, in particular for the national income accounts (NIA) estimates that will come out later this month. Examining these consequences is important, because the new data have the potential to affect our assessment of the macroeconomy, and thereby our view of the appropriate macro policy stance.
Start with the implications for the NIA.
1. The discrepancy between NIA and IIP manufacturing data will narrow: Ever since the new NIA methodology was adopted there has been a puzzling divergence between the robust NIA estimates of manufacturing performance and the weaker IIP data. For example, the NIA for 2016-17 has estimated real manufacturing growth at 7.7% while the (old) IIP had suggested zero growth.
The methodological sources of the discrepancy are well-known. The NIA estimates are based primarily on the financial accounts of companies while the IIP data are based mainly on volume data obtained from establishments. What has been less clear is why the two measures paint such different pictures, and which one is the better measure of manufacturing performance. Now, the pictures have come closer into alignment, since IIP manufacturing growth for 2016-17 is now at 5%.
2. Real gross value added (GVA) estimates for 2016-17 will be revised upward. (Ceteris paribus, that is, assuming that the underlying data are not going to be significantly different from the second advance estimates presented in February).
■ GVA manufacturing growth will increase because of the upward revision in IIP growth. Recall that the NIA proxies unorganized sector growth (weight of about 22% in overall manufacturing) by the IIP manufacturing index (suitably re-based to reflect the informal sector product composition). An upward revision of 5 percentage points in IIP could consequently increase GVA manufacturing growth by over 1 percentage point and hence overall GDP growth by about 0.2 percentage points.
■ GVA real manufacturing growth will increase further because the WPI deflators are lower. The exact magnitudes will have to be worked out. But broadly, since WPI manufacturing inflation is now lower by about 1.2 percentage points, real manufacturing growth will be revised upwards on this account.
■ These two points will (unfortunately) have the effect of re-widening the discrepancy between the NIA and IIP estimates of real manufacturing growth.
■ On services, there is a subtle but no less important point. In the NIA estimates, some part of GVA growth in services (accounting for about 20% of total GVA) is estimated by deflating nominal GVA growth in services by the overall WPI deflator (a detailed analysis of this issue is provided in the Mid-Year Economic Analysis of December 2015). Now this deflator has been revised downwards by 2 percentage points. Correspondingly, a large proportion of real services, and hence real GVA, growth will be revised upwards.
3. The trajectory of economic growth will improve. Current NIA estimates suggest that aggregate real GVA growth was 7.8% in 2015-16 and 6.7% in 2016-17, indicating a deceleration in growth. The upward revisions to 2016-17, however, will be much bigger than those for 2015-16 since the discrepancies between the old and new IIP and WPI series are particularly large in 2016-17. This means that the new GVA estimates will suggest either a slower deceleration between 2015-16 and 2016-17 or perhaps even an acceleration.
It is important to think ahead and consider how we should view this possible new growth trajectory. One could put the issue as follows: given that the new IIP and WPI data are an improvement on the old, does it mean that the consequential changes to the GVA estimates are also an improvement? The answer to that would be unambiguously yes in normal times, but not alas for 2016-17 both in absolute terms and relative to 2015-16.
■ The upward revision to GVA manufacturing arising from the use of (the higher) IIP as a proxy for the unorganized sector may raise questions, since the informal sector experienced unique shocks in 2016-17. The old IIP with weaker growth may have been a more realistic proxy for informal sector performance in 2016-17.
■ Similarly, the upward revision to real services GVA arising from the lower WPI raises questions. The downward revisions to WPI opens up a renewed divergence with consumer price index (CPI) services of almost three percentage points in 2016-17. If CPI services is actually a better measure of service price inflation, then the use of the new lower WPI as the deflator for nominal service activity will overstate real services GVA growth.
If this assessment is plausible, then macroeconomic analysis should also give serious weight to the new, much-improved IIP to complement a cautious reliance on NIA estimates. Three key implications follow from this analysis:
■ In 2016-17, formal manufacturing grew slightly slower than suggested by the NIA but at a still impressive rate given the weak state of the world economy.
■ But since the middle of last year there has been a noticeable deceleration of manufacturing activity.
■ In light of the above, there is a strong case for broad macro policy support, including monetary policy support, to reinvigorate the economy. There may also be a case for preventing a further deterioration in our exchange rate competitiveness.
Arvind Subramanian is chief economic adviser to the Union ministry of finance.