Pooja Mirchandani: A mostly normal monsoon

Although rainfall has been below normal in some parts of the country - the floods notwithstanding - the overall impact on major crops has not been too significant.

Although aggregate precipitation until August 16 was only 2 per cent below normal, the monsoons this year have been somewhat erratic in their temporal and spatial distribution. After setting in early during the last week of May, the rainfall almost dried up in most parts of the country during June. Although it improved to some extent during July, however, till July 26 too, the overall rainfall in the country was still 14 per cent below normal. Aggregate rainfall situation has, however, improved significantly thereafter. The cumulative rainfall till August 2 was just about 7 per cent below normal.

This was, however, owing to excessive rainfall witnessed in some parts of the country during end of July and beginning of August. Gujarat, Maharashtra, Rajasthan, Madhya Pradesh, Chhattisgarh and Orissa were the states that witnessed heavy rainfall during the week ended August 2. The trend of overall excess rainfall continued during the subsequent weeks also. And it took the cumulative rainfall during the current monsoon season to just 2 per cent below normal.

Aggregating the meteorological sub-divisions into regions, the country's western region had seen a spurt in rainfall activity during August. From a deficiency of 11.2 per cent till July 26, rainfall during the current monsoon went up to excess of 20 per cent till August 16. On the contrary, eastern, northern and southern regions witnessed a deficient rainfall of 22, 20 and 10 per cent, respectively. All this implies that even though aggregate rainfall throughout the country is just 2 per cent below normal, the rainfall pattern is spatially skewed.

The southwest monsoon rainfall still plays a very crucial role in the country's agricultural performance. With only about 40 per cent of the total cultivable land under irrigation, such dependence on rainfall is understandable. However, besides the total quantum of rainfall, its spatial and temporal distribution is a significant factor in determining output. The northern region, with more than 75 per cent of land under irrigation, is relatively less vulnerable to rainfall compared to the other regions. On the other hand, the western region, with just about 27 per cent land under irrigation, is the most vulnerable to the vagaries of the monsoons. It has rained in excess in the western region during the current monsoon season, which might be bad for the kharif crop, but it will provide sub-soil moisture to the land, which can improve the prospects for the rabi crop.

Regarding temporal distribution, our analysis of the past data shows that from an agricultural production perspective, rainfall in July is the most critical, followed by August. The skewed pattern of the southwest monsoons has, therefore, created some uncertainty over the kharif output. However, the sowing pattern presents a better picture. Compared to last fiscal, most crops had shown higher sowing until August 21 this year. Only coarse cereals and groundnut have reported relatively lower sowing compared to last year. Groundnut sowing is lower by 19 per cent and sowing of coarse cereals by 2 per cent.

To measure the impact of rainfall deficiency, we use an index called Deficient Rainfall Impact Parameter (DRIP), developed by Crisil in 2002-03. DRIP is computed as a product of percentage deviation of rainfall and percentage unirrigated area. DRIP captures the deficiency of rainfall (measured as deviation from normal) as well as the vulnerability of a region (measured as percentage unirrigated area). Within the range zero-100, a higher value of DRIP indicates greater adverse impact of rainfall deficiency. DRIP, however, does not measure the adverse impact of excessive rainfall.

In a state level aggregation, the DRIP values for all the crops produced in the state are combined using each crop's share of cultivated area. Based on the state-aggregated DRIP (see table), Tamil Nadu is the worst impacted state followed by Uttar Pradesh, Haryana and Madhya Pradesh. Gujarat, Maharashtra, Orissa and Rajasthan did not witness cumulative rainfall deficiency during the current monsoon season and as a result have DRIP values of zero. The drought year 2002 witnessed worst DRIP scores in past six years for all the states. However, DRIP value for Tamil Nadu during 2006 is higher than that in 2002.

To measure crop-aggregated DRIP (see chart), production levels of the crop in each of the major producing states are used as weights in combining the state-specific DRIP values for the crop. Based on crop-aggregated DRIP values, groundnut is the worst-affected crop, which is indicated by its sowing performance as well.

There exists a strong negative relation between Crisil's DRIP and agricultural GDP growth rate during past years, indicating that DRIP can predict agricultural output. Although we will come out with a formal agricultural forecast once full information on monsoons, that is, from June to September, is available, the indications from this exercise are that things should not deviate too much from "normality". Although rainfall has been below normal in some parts of the country (the floods notwithstanding), the overall impact on major crops, as our DRIP scores suggest, has not been too significant. July and August turned out to be a relatively wet months; our model linking DRIP to agricultural outcomes assigns a high weight to this.

The author is economist, CRISIL