I remembered that Sajjid Chinoy of JP Morgan (India) had written a piece on inflation dynamics in India in MINT last month. I went back and caught up with it. That reminded me of his piece last year. I read that too. Both are important articles and are well written. What follows is more of an analysis than criticism. Please bear that in mind as you read it.
Econometric analysis confirms the importance of Minimum Support Prices (MSP). After controlling for other potential determinants of cereals inflation, MSPs have a large and statistically significant impact on inflation, with a 1 percentage point increase in cereal MSPs increasing cereals inflation by 0.5 percentage points.
But what’s not ordinarily evident or intuitive is the role of global food prices. When we add them to the mix, we find that they have a large and significant impact on cereals inflation, to the point that the relationship between MSPs and cereals inflation becomes statistically insignificant. What this suggests is that the variation in MSPs is largely mimicking that of global food prices. [Emphasis mine]
… While there was always a suspicion that global food prices mattered in MSP-setting, its empirical strength is striking….
…our key finding is that the influence of global food prices—both directly and indirectly—on India’s food prices is larger than is suspected….
… a secular decline of global food inflation over the last year has likely had an important impact on India’s food disinflation…
…. What this suggests is that the sharp gyrations in input prices—led by global commodities—have been the key drivers in shaping core momentum in recent months, even as output gaps are slowly closing….
In sum, he finds that global factors – global food (impacts Indian MSP) and global commodities prices (impacts Indian Input prices and hence core inflation)
(2) Now, let us read his September 2016 article, which is a very highly accessible summary (no mean feat) of his joint work with two other authors:
In this MINT article, he makes two points. One is the average tendency based on fifteen years of data and the second one is on the recent inflation moderation. There is a bit of contradiction between the ‘average tendency’ and recent inflation drivers. There is nothing wrong about it. Specific episodes may deviate from average tendency.
His findings on the ‘average tendency’ (long-run inflation drivers in India):
(a) Output gap; (b) Past inflation; (c) MSP matters less once past inflation is taken into account (this contradicts last year’s note). If it means that MSP is set based on last year’s inflation, then that begs the question of what drove last year’s inflation and so on; (d) Regime change in dummy – he concedes that they are not quite clear as to which regime change it is actually trying to capture and (e) Oil prices and rainfall shortage (if shortfall is more than 15%).
From the above ‘average’ findings, decomposition of recent inflation drivers:
(i) 40% adaptive expectations; (ii) 20% to reduction in MSP growth; (iii) 30% to New Regime and Commodity Prices.
But, he gives room for head-scratching by stating the following:
Interestingly, and contrary to popular perception, the actual fall in global oil prices (apart from any impact on inflation expectations) explains very little of the disinflation, because such a small fraction was passed on to households.
In general, regression estimates capture ‘average’ tendency and a specific time period may or may not conform to the regression estimates and findings.
Therefore, if we ignore his ‘decomposition’ findings for now and focus on the ‘average’ tendency and compare it to what he wrote last year, he is contradicting his 2015 article in MINT on the role of MSP (and, in effect, global food prices).
So, do global food prices matter or not, for inflation generation process in India?
IMF ‘Selected Issues’ documents for India issued in 2014 and 2015 seem to suggest that they do matter and that is in line with what Sajjid wrote in 2015 for MINT.
He is consistent on non-food commodities though. Between 2015 and 2016 ‘average tendency’, he states that global commodities matter.
Only in decomposing the more recent inflation deceleration, he finds that oil price did not play a big role. That is NOT quite inconsistent. Oil price may not have influenced recent inflation trends in India (in the last one year) because the end price for consumers did not change much, due to the government’s revenue mobilisation efforts via additional taxes that neutralised the drop in the international price of crude oil.
I think he leaves us with four factors: Output gap, Global Food Prices, Past inflation and domestic Input prices (which are a function of global commodities prices).
Where he has caused us confusion is in the role of global food prices directly or via MSP. That was strong in 2015 but he has diluted it in 2016. Second, past inflation is really no guide. We know that it matters because inflation rates are sticky and so are inflation expectations. But, that leaves us with unanswered questions: what drove last year’s inflation and so on?
Vivek Dehijia has a good summary of these works and other papers here.