The quality problem in education and health care

The issue of quality in public service delivery has been a constant theme in this blog. This post covers the latest evidence on this front from school education and basic health care in India. 

Alexis Le Nestour, Laura Moscoviz, and Justin Sandefur have a working paper which examines quality and access in school education across 87 countries in the 1950-2000 period. Its findings are disturbing. Lant Pritchett has a blog here. It has this about India's learning outcomes progress,

In India between the birth cohorts of 1958 and 1995 those with 5 or more years of schooling expanded from 36.6 to 77.3 percent. But the likelihood a woman completing just primary school could read, Read@G5 fell from 89.8 percent to 51.4 percent.

In contrast, successes of Peru and Vietnam,

Peru shows a positive picture in both dimensions. In the birth cohort of 1952 71.7 percent of women completed Grade 5 or higher and that rose to 93.7 by the birth cohort of 1992. Similarly, the share of those with Grade 5 complete who could read rose from 69.4 percent to 84.9 percent. In Vietnam between the birth cohort of 1958 and 1994 the fraction with schooling of grade 5 or more increased from 76.7 to 88.7 and Read@G5 rose from 72 to 94.1 percent.

This graphic pretty much sums up the general trend of declining quality - India's performance is among the poorest, worse than any of its neighbours. Pakistan and Indonesia improved their learning quality.

Quality is a problem in health care too. A friend pointed me to a Gates Foundation commissioned Harvard TH Chan School study to assess health care quality in Odisha. They used the now standard approach of vignettes study - send decoy patients ostensibly suffering from various diseases, and measure the diagnosis done by doctors (questions asked as per the WHO treatment protocols), average time spent, and accuracy of diagnosis. 

Their findings are here (see slides 15-18) and here. Just 6.8% of providers (MBBS doctors) diagnosed TB! A summary,

No significant difference between MBBS doctors, AYUSH & unqualified. Providers trained at government colleges were slightly more competent than those at private institutes. Providers in urban areas were more competent than those in rural areas. Private sector providers were more competent than public sector providers at PHCs. There was no significant difference between providers who received inservice training versus those who did not. Diagnostic competence was not significantly correlated with the amount of time spent per patient. No significant difference between providers with more or less years of work experience...

~60% cases diagnosed correctly. Providers wrongly diagnosed as a less serious illness (E.g., cold, fever for TB, headache for preeclampsia, acidity & body ache for heart attack). Only ~2% providers advised correct treatment. Although ~50% prescribed at least one correct drug. ~40% prescribed only unnecessary (sometimes harmful) drugs/antibiotics – raising concerns of low-value care, anti-microbial resistance.

This is confirmed by several similar vignettes studies in recent years - this one even finds Indian physicians (in Delhi and MP) being worse than physicians from Tanzania and Indonesia in the diagnosis of multiple medical conditions. 

In this context, it strikes me that there is little public acknowledgement of the health quality problem, even though it's just as dismal as learning outcomes which is now widely acknowledged. Perhaps this highlights the importance of agencies like Pratham and its ASER reports in highlighting the issue. I don't know of any such large and long-drawn non-government efforts that focus on health care quality. 

We now have a major problem in the quality of health and education being delivered. There are at least two factors at work. One, at the intensive margin, for a variety of factors, the average quality of educational instruction and health treatment is itself declining. Two, at the extensive margin, the rapid expansion in supply of training institutions (medical and degree colleges, and teacher training institutes) and professionals (teachers, doctors, paramedics etc) is naturally eroding the average quality of professionals passing out from these institutions. This is re-affirmed by the findings of the Sandefur et al paper - "the fall in observed quality is greater where enrollment grew faster, and after the abolition of user fees". 

One policy point to think about. How about a five year phased increase in allocations under the flagship National Health Mission (NHM) and Samagra Siksha Abhiyan (SSA) towards quality (and away from infrastructure and inputs)? Let the Health Department of Government of India do something for assessing the quality of primary and secondary health care which is similar to NAS (National Assessment Survey) which measures learning outcomes in Grades 3, 5, and 8. 

Ideally, there should be independent standardised quality determination through sample surveys done by DMEO or privately (and not by the Department). The Health Department in consultation with DMEO should formulate survey instruments and survey methodology. It should then initiate periodic standardised quality surveys through independent agencies across the country. A combination of this, coupled with quality focused inputs and processes, should form the focus of fiscal resource allocations.