ISSN (Online) - 2349-8846
-A A +A

Impediments to the Spread of Crop Insurance in India

Subhankar Mukherjee (subhankarm13@iimcal.ac.in) and Parthapratim Pal (parthapal@iimcal.ac.in) are at the Indian Institute of Management Calcutta, Kolkata.

The Government of India aims to double the crop insurance coverage to 50% through the Pradhan Mantri Fasal Bima Yojana by 2018. By analysing the agricultural household data, this article comments on the feasibility of attaining this target by looking at the past performance of similar schemes. Few critical areas where efforts need to be concentrated in order to increase coverage are discussed.

The authors thank Santanu Bhadra of IIM Calcutta for valuable review comments. 

On 13 January 2016, the central government announced a new crop insurance scheme for farmers in India. The scheme, named Pradhan Mantri Fasal Bima Yojana (PMFBY) is going to complete one year of its launch soon. One of the major emphasis of the scheme has been to double the coverage of crop insurance among farmers to 50% by 2018.1 On 7 December 2016, the government announced that it expects the coverage to reach to over 30% farmers by the end of 2017 (Sally 2016). This article analyses the coverage of crop insurance schemes in India using household level data. It also comments on the feasibility of attaining 50% coverage by looking at the past performance of similar schemes. Further, this article points out a few critical areas where efforts need to be concentrated in order to increase coverage.

Crop Insurance

The proposition of insuring crops in India dates back to 1920 (Mishra 1995). Since independence, efforts were made both at the central as well as the state level to introduce crop insurance schemes for Indian farmers. In 1972, the first crop insurance scheme was launched. The latest was launched in 2016. Raju and Chand (2007, 2008) provide a detailed account of the evolution of the crop insurance schemes in India. We provide a sketch of this evolution in Table 1.

Designing a cost-effective crop insurance scheme has proved to be challenging because (i) the risk of crop loss has a significant systemic component; and (ii) ex ante risk assessment and ex post loss assessment for individual farmers are costly. Dandekar (1976) proposed an area index-based approach as a second-best solution to reduce these costs. Under this approach, crop loss is recognised when the average yield of a large area is lower than a pre-specified threshold yield. Weather index-based crop insurance is another approach to crop insurance, wherein crop loss is recognised when rainfall in an area is lower/higher than pre-specified threshold-level rainfalls.

In India, crop insurance schemes are mandatory for farmers who are taking short-term crop loan (loanee farmers). But it is voluntary for all other farmers (non-loanee farmers). The premium amount is deducted from the loan amount for loanee farmers, and claim, if any, is adjusted against the loan amount. The non-loanee farmers, pay premium from their pocket, and receive the claim amount if crop loss is recognised.

Data Sources

Two separate data sets have been used for this analysis. The first set of data is from National Sample Survey Office’s (NSSO) Situation Assessment Survey of Agricultural Households, 2013. This survey was conducted in two visits to the same set of households for the agricultural year 2012–13. In the first visit, information was collected for the period July to December 2012 (hereafter referred to as period 1), and in the second visit, data was collected for similar set of questions for the period January to June 2013 (hereafter referred to as period 2). The sample size of the survey was 35,200 agricultural households spanned across 36 states and union territories.

The second set of data used in this analysis is collected from the Agriculture Insurance Company of India’s (AICI) Business Profile Data, available on the company website.2 This data set provides information on crop insurance adoption in India for states and union territories for the period 2000–13, separately for two cropping seasons and for three major crop insurance schemes, National Agricultural Insurance Scheme (NAIS), Modified NAIS (MNAIS) and Weather Based Crop Insurance Scheme (WBCIS).

Coverage of Crop Insurance

There is no standard method to compute the coverage level of crop insurance in India. Coverage can be based on the proportion of farmers insured, or on the proportion of area insured. This distinction is important because of the huge inequality in landholding (Rawal 2008). Also, crop insurance is sold separately for each crop and for each of the two cropping seasons. Evidently, a farmer may buy more than one insurance in one cropping season, and also buy insurance in both cropping seasons. According to NSSO data, the farmers who bought insurance, bought 1.45 units of insurance in period 1 and 1.64 units in period 2, and 30% of farmers who bought insurance in period 1 also bought insurance in period 2.

We estimate the coverage of crop insurance by (i) number of farmers; (ii) units of cultivation; and (iii) area insured. As discussed, a simple addition of the number of farmers who bought crop insurance in both the survey periods would yield an (upwardly) biased result. So we subtract the number of overlapping farmers from the sum and then divide the resultant by the number of farm households surveyed to arrive at the coverage level by number of farmers. The approach is depicted in equation (1).

... (1)

 

where,

fs= Percentage of farmers covered under crop insurance in states;

fKs= No of farmers covered under crop insurance in kharif season in states;

fRs= No of farmers covered under crop insurance in rabi season in states;

fBs= No of farmers covered under crop insurance in both seasons in states;

fSs= No of farmers surveyed in states.

In the second approach, we compute coverage of crop insurance by units of cultivation insured. We define one unit of cultivation as one crop sown by an agricultural household at any time during an agricultural year.

Table 2 shows the coverage percentages estimated. The results have been tabulated for the top 14 states according to the share of number of farmers insuring in a year (as obtained from the NSSO data). About 94% of insured farmers and 95% of insured units of cultivation are from these 14 states. Figures at the all-India level are, however, based on data from all 36 states and union territories.

Table 2 reveals that coverage of crop insurance in India is abysmally low both by number of farmers and units of cultivation. Annually, just over 7% farmers subscribe to crop insurance (column a), and only 4% of the units cultivated are insured (column d). These figures are particularly striking because crop insurance in India is mandatory for all farmers taking short-term crop loans. Banks are supposed to deduct the premium amount while advancing crop loans to farmers. Given the disparity between crop loan penetration and crop insurance penetration, it is apparent that this rule is not strictly followed in many states. Crop insurance is also available on voluntary basis for non-loanee farmers. But, as we discuss later, voluntary purchase of crop insurance in India is also extremely low. Seasonal coverage data are presented in columns b, c, e and f. Adoption rate for period 2 is lower, since crops sowed in that period are less dependent on rainfall, and thus, less risky.

Distribution of coverage among states varies widely. While annual coverage by number of farmers (column a) in Andhra Pradesh and Chhattisgarh is more than 20%, the same for Bihar and Uttar Pradesh is around 3%. Rest of the states and union territories, not included in Table 2 have, naturally, much lower penetration rate.

In the third approach, we estimate the proportion of area insured under crop insurance. We add area insured for all crop insurance schemes for both the seasons and then divide the sum by total cropped area—obtained from Land Use Statistics, 2012–13—to obtain this coverage level. The figures are tabulated for the 14 states in column g of Table 2. These states constitute of 98% of area insured. As shown, around 19% of cultivable area was insured in 2012. We cannot estimate season-wise coverage of area due to unavailability of data for total cropped area. A comparison of the coverage of crop insurance figures in columns a, d, and g of Table 2 indicates that crop insurance adoption is disproportionately tilted towards large landholding farmers. Table 3 corroborates our intuition. Proportion of farmers buying crop insurance increases monotonically from marginal to large landholding farmers in both periods.

Growth Rate of Crop Insurance

NSSO data does not allow us to assess the growth rate of coverage, since the survey was conducted for one year. So, we have used AICI data, combined for all schemes from 2001 to 2013 (since data is available for this period for most states), to measure the average growth rate in crop insurance take up by number of farmers for each season as well as annually. We tabulate the result in Table 4 for the same set of states for which the coverage level was calculated, except for Telangana, since data for the newly formed state is not available.3

Average growth rate in terms of number of farmers for the 12-year period is 6.5% annually. A closer look at the year-wise trend in growth rate reveals that there were sharp increase in growth rate in 2004 and in 2009, and since then, the rate has slowed down considerably (Figure 1). Growth rate of crop insurance in the kharif season is lower than that for the rabi season, which can be explained through lower base in the latter.

Like coverage, growth rate of crop insurance also varies substantially among states. In Table 5, we arrange the states according to the annual coverage level by number of farmers and average annual growth rate. We consider an annual coverage level of 10% or more as high coverage, and an annual growth rate of 7% or more as high growth rate. As shown, Kerala, Maharashtra, and West Bengal will need special attention in order to improve overall coverage and growth of crop insurance penetration.

Reasons for Low Demand

Cole et al (2012) and Stein (2011) provide evidence to show that in India, demand for crop insurance is highly price-sensitive and depends upon prior experience. The present crop insurance scheme (PMFBY) has lowered the premium rates to 1.5% of sum insured for kharif crops, 2% for rabi crops and 5% for commercial crops. This step may attract some more farmers, but NSSO data suggests that around 60% farmers were either “not aware” or “not aware about availability of facility” to buy crop insurance. So, policies to increase awareness need to be pushed to increase coverage further.

Another reason for low confidence on crop insurance is the delay in payment of claims. The reason for delayed payment is due to lengthy procedure in gauging the extent of crop loss (Nair 2010). In the NSSO survey, non-loanee farmers were asked about payment of claim amount when they faced crop loss. Figure 2 shows the distribution of these responses. For an overwhelming proportion of units of cultivations, farmers did not receive any claim amount even when they faced crop loss. Also, as crop insurance schemes are indexed schemes, receipt of claim does not depend on the farmer’s individual loss, but rather on the average loss of a region. The smaller the size of the area of the region, the more the accuracy of estimation of crop loss for an individual farmer. The MNAIS, launched in 2010, tried to address these issues through initiation of partial on-account payment and reducing the area size wherever feasible. PMFBY has continued these features, but implementation will remain an area of concern.

Reasons for Low Supply

Even though crop insurance is mandatory for loanee farmers, there is a huge gap between adoption rates of crop loans and crop insurance. This may be the result of poor enforcement of the rule of compulsorily insuring loanee farmers. Further, the proportion of farmers buying insurance without taking loan is only around 15% of the total farmers buying crop insurance (Table 6). Crop insurance is sold at bank branches in most parts of India. So, poor adoption rate by non-loanee farmers can be attributed to lack of incentive for banks to sell crop insurance to farmers who do not subscribe for loans.4 Initiation of other avenues such as outlets to sell only crop insurance will increase take up.

Our analysis of NSSO data shows that incidence of crop loss reported is higher when crops are insured than when not insured (Table 7). The problem is more acute when crops are insured voluntarily (that is, without taking crop loan). Incidence of such loss is high in Andhra Pradesh, Gujarat, and Rajasthan. This may be due to two reasons. First, due to the presence of well-recognised problem of information asymmetry in supply of insurance contracts (Newbery and Stiglitz 1981), farmers do not have adequate incentive to avoid risk once crops are insured (moral hazard problem), and the farmers who anticipate a crop loss will be more inclined to buy insurance (adverse selection problem). Second, the cut-off date to apply for crop insurance is extended close to the harvesting time in some seasons. In 2016, the cut-off date for kharif season was extended to 10 August. This allows farmers additional time to procure insurance if they foresee crop loss.

Conclusions

According to our analysis, 7.22% farmers were covered under crop insurance in 2012–13 and the average growth rate of crop insurance adoption from 2001 to 2013 was 6.48%. With this level of coverage and growth rate, less than 10% farmers would be covered in 2016–17. Thus, attaining a coverage of 50% of farmers will take a long time. According to the government, however, 23% farmers were covered under crop insurance in 2015–16 (Indian Express 2016), and around 26% farmers have been covered in 2017 so far.5

As this article highlighted, existing evidence suggests that the key issues of low spread of crop insurance are lack of awareness among farmers, delay in claim settlement, absence of adequate number of channels, and lack of information on the risk behaviour of farmers. The PMFBY has brought about lower and standardised premium rates, and emphasised the use of technology.6 However, as discussed, there are some other structural constraints that may need to be tackled if the targeted coverage of crop insurance in India needs to be achieved.

Notes

1 See, http://pib.nic.in/newsite/PrintRelease.aspx?relid=134432, and Indian Express (2016).

2 http://www.aicofindia.com/AICEng/Pages/MapOfIndia_BP.aspx.

3 Some states show exceptionally high or low growth rates for some years. For example, growth rate in Rajasthan for 2004 is over 2000%. We have excluded such outlier values from our calculation.

4 Planning Commission mentioned that crop insurance helps banks more than farmers (Twelfth Five Year Plan document).

5 Rajya Sabha unstarred question number 322, dated 18 November 2016, shows the total number of farmers insured during kharif 2016. We have divided this by the total number of cultivators obtained from Census 2011 to arrive at this figure.

6 As per Lok Sabha unstarred question number 263, dated 19 July 2016, the government is developing drone-based crop and soil health monitoring system, which could be integrated with satellite-based technologies in future.

References

Cole, S et al (2012): “Barriers to Household Risk Management: Evidence from India,” Working Paper WP/12/95, International Monetary Fund.

Dandekar, V M (1976): “Crop Insurance in India,” Economic & Political Weekly, Vol 11, No 26, pp A61–A80.

Indian Express (2016): “Govt Aims to Bring 50% Farmers under PMFBY in Next Few Years,” 15 March.

Mishra, P K (1995): “Is Rainfall Insurance a New Idea? Pioneering Work Revisited,” Economic & Political Weekly, Vol 30, No 25, pp 84–88.

Nair, R (2010): “Crop Insurance in India: Changes and Challenges,” Economic & Political Weekly, Vol 45, No 6, pp 19–22.

Newbery, D and J Stiglitz (1981): The Theory of Commodity Price Stabilization, Oxford University Press.

Raju, S S and R Chand (2007): “Progress and Problems in Agricultural Insurance,” Economic & Political Weekly, Vol 42, No 21, pp 1905–08.

— (2008): “Agricultural Insurance in India: Problems and Prospects,” Working Paper No 8, National Centre for Agricultural Economics and Policy Research, Indian Council of Agricultural Research.

Rawal, V (2008): “Ownership Holdings of Land in Rural India: Putting the Record Straight,” Economic & Political Weekly, Vol 43, No 10, pp 43–47.

Sally, M (2016): “26.5% Farmers Sought Coverage under PM’s Crop Insurance Scheme,” Economic Times, 7 December.

Stein, D (2011): “Paying Premiums with the Insurer’s Money: How Loss Aversion Drives Dynamic Insurance Decisions,” Working Paper, Agricultural and Applied Economics, https://aae.wisc.edu/mwiedc/papers/2011/Daniel_Stein.pdf.

Updated On : 6th Sep, 2017

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top