Graphic: Subrata Jana/Mint
Graphic: Subrata Jana/Mint

Rains, reservoirs and food prices

Monsoons don't impact food inflation directly, but through their association with reservoir levels

India’s summer is laden with suspense. Several matters are up in the air. Will oil prices continue their ascent? After the unpleasant surprise in April, have we entered a period of rising inflation? Will we get good rains that keep food prices in control? Or, taking a step back, do rains really matter for food prices?

Only time can provide the answer to most of these questions. But there is one question that the past can shed some light on. We looked at what happened over 15 years and found there is no simple correlation betw-een the monsoon season (June-September) and food prices. What matters is not the rain, but the reservoirs.

We analysed daily data on the water stored in 90 of the country’s largest reservoirs to see if they were associated with food prices. The advantage of using reservoir data is that not only do they tell us about rain that has just fallen, but also rain that was stored, especially from non-seasonal showers, which are becoming more frequent with climate change. They also serve as a sort of proxy measure of groundwater levels. Reservoirs tell us not just about how much water has been stored, but also how much moisture there is in the soil when it is needed. Finally, they also reflect cost of production directly, given that surface water irrigation (via reservoirs) is cheaper than groundwater irrigation (via mechanical pumps).

The data partly helps explain irregular years such as 2015 when monsoon rains were deficient (14% below normal) and yet, food inflation remained contained. Perhaps it was the sufficient water stock in India’s reservoirs (6% above normal in the pre-monsoon month of May), thanks to unseasonal rains from a few months before that helped keep food inflation low. Or years such as 2008, 2010 and 2012 when monsoon rains were normal, but reservoir levels were in deficit and food inflation accelerated.

In particular, we found that reservoir levels in July are most significant. Not surprising, because June and July are the important sowing months (see chart 1). We also found that July reservoir levels are determined by a combination of May reservoir levels (that indicate the carry-over from pre-monsoon months, especially the remains from prior rain spells), and June and July showers.

Since we wanted to know the overall impact on food price inflation, we also looked at other factors that may drive it. Of the many we considered, minimum support prices (MSPs) in agriculture and global food prices turned out to be important. We found that MSPs, in turn, are explained by rural wages (domestic factors) and crude oil (international factors), given that the two double up as proxies for all other input costs. But we also know that setting MSPs entails a high level of discretion. As a result, the explanatory power of these relationships is limited. (For the statistically inclined, the r-squared of the regression is no higher than 35%).

It will come as no surprise that global food prices are also important in explaining inflation in India, especially in recent years, when agricultural trade has intensified, increasing the links between domestic and global prices.

We now have three ‘variables’ affecting food prices—MSPs, global commodity prices and reservoir levels. Our ‘model’ works well in explaining past food inflation episodes (see chart 2).

There was one more thing we wanted to be sure of. It’s quite possible that reservoir levels matter (in driving up inflation) when they are low, but they do not matter (in driving down inflation) when they are high. To test this, we converted our July reservoir indicator into two ‘dummy variables’—one that switched on when reservoir levels fell and the other switched on when reservoir levels rose. We found that reservoir levels have the same impact on the way up as they do on the way down.

Putting it all together and applying to current conditions, we find an interesting tug-of-war at play. Of the three drivers of food inflation, two are at odds with each other. Global commodity prices are rising (both fuel and some food products like sugar), but MSPs have been kept in check (under 5% year-on-year for the key produce such as paddy). That leaves the last candidate, reservoir levels, to determine food prices for the remainder of the year.

And because this year we are entering the season with a notable deficit in reservoirs (water levels were at 17% of capacity at the end of May versus 27% at the same time last year), early rains will matter much more than before.

In short, rains may not impact food prices directly, but through their association with reservoir levels, especially this year, they are likely to. Sufficient and timely rains are now critical. In fact, every drop will count.

Pranjul Bhandari is chief India economist, HSBC.

Comments are welcome at theirview@livemint.com

Close