Economists examined the national suicide data for 1997–2007 and concluded that there was no link of farmer suicides to Bt cotton adoption
19]. Indeed, plots of annual suicides in GJ and KA show no trend with time (Fig.
10a) or on the national total of suicides. Suicides in GJ were about 500 per year (<3 % of the national total), while cotton production was 24–39 % of the national total
37], and average yields were mostly >500-kg lint cotton ha −1 with CV <50 % (Fig.
7b). Farmer suicide rates in KA were high at about 2000 annually, but cotton production was only 2–4 % of the national total with predicted yields <500 kg and CV >50 % being very common (Fig.
7c). In KA, these are indicators of risk not only in cotton but also in other parts of the agriculture sector due to low and highly variable rainfall. In AP and MH, suicides are strongly increasing with time (Fig.
10b) and on the national total (Fig.
10c) with the increase beginning
before the introduction of Bt cotton in 2002.