While the aggregate Sustainable Development Goals (SDG) seem realisable, the disturbing part of the story is the stark intra-country disparities that reveal differential vulnerabilities across regions.
With the commitment towards compliance with the SDGs, there is a definite improvement in the data system to obtain the relevant estimates for indicators representing each of the goals. In the process of obtaining these estimates, two concerns arise: the levels of disaggregation and the reliability of the estimates conditioned by the rarity of the event or phenomenon.
In terms of an adverse indicator, like the maternal mortality rate,, the two concerns are equally relevant in the sense that with decline in this adversity, there is a compromise in its reliability. Further, its disaggregation obtains differentially reliable estimates ineligible for comparison.
A 2015 WHO study on the reduction of MMR by 75% from 1990 to 2015 in the Millennium Development Goals (MGDs) suggested that, by 2013, MMR was reduced by 45%, from 380 deaths per 100,000 live births in 1990 to 210 deaths per 100,000 live births – reflecting substantial progress but still falling far short of the global MDG goals. MMR reduction is a priority in the SDGs, with the broad objective to reduce the global maternal mortality ratio to less than 70 per 100,000 live births by 2030.
Hence, it is questionable as to whether such estimates be obtained in brief periodicity (i.e. every three years) based on the same sample coverage and secondly should these estimates be compared across regions with differential reliability. The following discussion elaborates on this pressing concern to draw attention of policy makers.
The most recently released statistics on maternal mortality notes a decline of 8 units in the maternal mortality ratio in the gap of three years, which is a positive trend. The target level of SDGs for this indicator is 70 per one lakh live births by the year 2030, and this seems very much within the horizon given the declining exposure to reproduction on one hand and ever-increasing extent of institutional deliveries on the other. While this aggregate achievement seems realisable, the disturbing part of the story lies in the intra country disparities which are stark and revealing of the differential vulnerabilities across regions.
The national achievement makes a specific region 1.8 times more vulnerable vis-à-vis the other two regions 0.52 and 0.63 times-less vulnerable based on the MMR. These differences are alarming in the sense that they are not merely guided by the differential risk of reproduction alone as convergence in fertility decline across the country is a reality. Further, such a pattern could very well lead to compliance with the SDG target at the national level leaving behind some regions with such adversities. The race towards compliance with SDGs has always this limitation as one of the authors of this note observed in the initial years of SDG formulation as to whether target compliance will compromise on equities.
The three regional estimates of MMR put the southern region comprising five states at 72, a group of other six states of the north and western region at 90 against the Empowered Action Group (EAG) states with the inclusion of Assam at 175. Apart from these regional estimates, Assam leads the table with an MMR levels of 229 followed by Uttar Pradesh, Madhya Pradesh and Bihar which have MMR estimates closer to 200. One would imagine that these are purely a derivative of differential exposure to reproduction but it is not the case given the decline in fertility levels across the EAG states.
There is apprehension as to such high levels of MMRs may not necessarily be associated with childbirth alone but miscarriages and non-clinical abortions as well. The levels of MMRs in EAG states are distinctly higher compared with other states except in Punjab. Another aspect of this comparison relates to the statistical reliability of these estimates. The high levels of MMR in EAG states do have the widest possible confidence intervals understandably owing to the rarity of this event. But the kind of decline that is celebrated should also take into account the wide confidence intervals as they might otherwise be a statistical construct and not a real decline at all. In that sense, the regional estimates are somewhat acceptable for comparison which presents a clear regional divide that needs focus and attention.
Another aspect of this phenomenon is to situate them in an age profile to identify age-specific vulnerabilities. Given that a larger share of fertility or childbirth related experience is below the age of 30 years, a risk evaluation indicates that almost two-thirds of maternal deaths are happening between ages 20-30 years of age. Hence, the common understanding that adolescent motherhood or high-risk age of maternity do not seem to intensify this problem. In addition, the fertility transition too places this age bracket to share a significant share of the childbirth experience. This suggests towards targeting the younger and primipara women to comply with institutional deliveries as well as the maternal care continuum to avoid such adversities.
Finally, caution needs to be exercised about certifying compliance of such targets overlooking the regional disparities as well as reliability of estimates given the rarity of the phenomenon itself. While monitoring this phenomenon is desirable, it is rather important to change focus from reducing levels to that of converging them.
Udaya S. Mishra is a professor of economics at the Centre for Development Studies, Thiruvananthapuram. Balakrushna Padhi works is an economist at the Centre of Excellence in Fiscal Policy and Taxation, Bhubaneswar. The views express here are personal.