Page 312 - ES 2020-21_Volume-1-2 [28-01-21]

JAY Ho: Ayushman Bharat's Jan Arogya Yojana (JAY) and Health Outcomes 295

9.21 It is interesting to note that pre-authorized volumes of claims >INR 2500 significantly
exceeded the pre-authorised volumes of claims < INR 2500 before CoVID-19. This gap
suggests a strong preference for costly procedures and tertiary care early into the adoption cycle
of PM-JAY up until the distruption caused by the CoVID-19 pandemic. After the CoVID-19
pandemic distruption, there is a reversal of this trend with the number of pre- authorised claims
which are <2500 INR exceeding the the number of pre-authorised claims claims >2500 INR.
This is indicative of an increase in the utilization of PM-JAY for non-costly procedures and PM-
JAY even being used as a substitute for primary care.

PM-JAY AND HEALTH OUTCOMES: DIFFERENCE-IN-DIFFERENCE
CALCULATIONS

9.22 In this section we explore whether access to the PM-JAY scheme has had any significant
impact on the health outcomes. We compare the health indicators measured by National Family
Health Survey-4 (NFHS 2015-16) and the National Family Health survey 5 (NFHS 2019-20)
to undertake this analysis. As PM-JAY was implemented in 2018, these two surveys provide
before-after data to assess the impact of PM-JAY with the NFHS-4 serving as the baseline to
compare the changes using NFHS-5. To mitigate the impact of various confounding factors,
including but not limited to secular improvements in health indicators across the country, we
undertake this analysis by estimating a difference-in-difference. The Economic Surveys of 2018-
19 and 2019-20 have discussed the econometric benefits of this technique to account for various
confounding factors and thereby assess the impact of a policy change on outcomes. We refer
readers to these surveys for technical details. In essence, we compute the before-after difference
in outcomes for a state or group of states that implemented PM-JAY and compare the same
before-after difference in a state or group of states that did not implement PM-JAY. The latter
difference provides an estimate for the counter-factual: what would have been the before-after
difference in outcomes for the state or group of states that implemented PM-JAY if they had not
implemented PM-JAY. Thus, by comparing the former difference with the latter difference, we
can reasonably attribute the difference-in-difference to be the impact of PM-JAY.
9.23 We undertake this analysis in two parts. In the first part, we use West Bengal as the state
that did not implement PM-JAY and compare the before-after difference in health outcomes to
its neighbouring states that have implemented PM-JAY – Bihar, Sikkim and Assam. Apart from
all these states being contiguous to each other and therefore being similar on socio-economic
dimensions, we show that the baseline characteristics of these two groups of states were similar.

9.24 In the second part, we repeat the same analysis for all states that did not implement PM-
JAY vis-à-vis all states that implemented PM-JAY. of course, the heterogeneity across the entire
group of states in the country is large. The second analysis is less of a like-for-like comparison
than the first one. Combining the findings from both these comparisons ensures that the findings
are robust not only to a more localised, and therefore, more careful comparison but also to a
comparison that spans all the major states in the country.

Comparing West Bengal versus its neighbours (Bihar, Sikkim, Assam)

9.25 We first compare West Bengal with its neighbours in key demographic and household
characteristics across the time span of NFHS 4 and NFHS 5. Figure 4 presents this comparison.