Tuesday, January 3, 2012

Lies, damned lies, and statistics: On Arvind Panagariya’s Kerala adventure

This is with reference to the article “Cracking the Kerala Myth” by Arvind Panagariya (Professor of Economics, Columbia University), that appeared in The Times of India, dated 2nd January 2012. It may eminently suit the political convictions of the Columbia Professor to criticise Kerala’s development experience from his free-market perspective; he is also free to argue that Kerala’s is not a “state-led success” and that its “left-of-centre governments” did not contribute to its success. However, while doing so, one expects a modicum of rigour in argument. Sadly, Panagariya’s piece falls flat, both in its historical grasp and statistical rigour.

First, Kerala’s high level of inequality in consumption expenditure, in absolute terms, is used by Panagariya to argue that the source of poverty decline was income growth, and not equitable development. He claims this to be true right from 1973-74. Panagariya is wrong. A simple plot of poverty data would have revealed to him that the sharpest fall of poverty in Kerala occurred from the early-1980s onwards. The most remarkable aspect of this poverty decline was that inequality levels also declined alongside. Thus, between 1983 and 1993-94, head count ratio (HCR) of poverty in rural Kerala fell from 39.6 per cent to 25.4 per cent. During the same period, the gini ratio, used to measure inequality, also fell from 0.32 to 0.30. The association holds for urban areas also, where, the gini ratio fell even sharply from 0.39 in 1983 to 0.34 in 1993-94 (in urban India as a whole, the gini ratio had risen between 1983 and 1993-94).

After 1993-94, poverty levels have continued to fall, while gini ratios have risen. Thus, between 1973-74 and 2004-05, there are two phases: Phase 1, where poverty and inequality fell together; and Phase 2, where poverty fell and inequality rose. Such contradictory outcomes do not allow the kind of generalisation that Panagariya attempts.

Panagariya also misses the point that till the late-1980s, Kerala’s economy was not on any growth path at all. Scholars describe the period between the early-1970s and the late-1980s as a period of “economic stagnation” in the State (the real NSDP grew at just 1.6 per cent per annum). It was only after 1987-88 that NSDP growth rates in Kerala begin to pick up, while poverty rates had begun to fall from the early-1980s itself.

Secondly, Panagariya makes the usual mistake that uninformed commentators on Kerala make. He argues that Kerala’s social indicators are good because “it started at the highest level at independence”. It is here that Panagariya stretches the science of statistics beyond imagination. He says: “In 1951, it had a literacy rate of 47% compared with 18% for India as a whole and 28% for Maharashtra, the closest rival among the large states. By 2011, these rates had risen to 94, 74 and 83%, respectively. The gains made, thus, equal 47, 56 and 55 percentage points for Kerala, India and Maharashtra, respectively.”

This use of statistics by Panagariya would shock even a Class 10 student of statistics. In his argument, if Kerala has to be considered as success, the absolute difference between the literacy rates of Kerala and India/Maharashtra should either increase or at least remain constant! Thus, Kerala should have had a literacy rate of at least 102 per cent in 2011 for Panagariya to call it a “success story”!

Similar is Panagariya’s use of statistics in the case of Infant Mortality Rate (IMR). In trying to argue that Gujarat achieved more in IMR than Kerala, he says: “whereas Kerala lowered its infant mortality rate by 46 deaths per 1,000 live births between 1971 and 2009, Gujarat achieved a reduction of 96, Tamil Nadu of 85 and Maharashtra of 74.” In 1971, Kerala’s IMR was 61 per 1000 live births. To be on par with Gujarat in 2009, Kerala should have had an IMR of -35!

Columbia’s statistics department can rejoice; in their university, the very basic principles of statistics are being rewritten! It is often said that there are two kinds of statistics, the kind you look up, and the kind you make up. No prizes in guessing where Panagariya’s statistics belongs.

But apart from statistical gaffes, Panagariya’s article also misses important historical features of Kerala’s development experience. One, it is important to appreciate that all social sector achievements of Kerala precede the late-1980s, when the State’s per capita income began to grow rapidly. Two, it is not enough to say that Kerala had higher literacy rates than other States in 1957. Here, absolute levels of literacy rates assume great significance. See Table 1 below.

The Kerala of 1956 was formed by merging the three provinces of Travancore, Cochin and Malabar. Malabar was certainly the most backward among the three. In 1951, the total literacy rate in Malabar was only 31 per cent, while it was only 22 per cent among women (Table 1). Among Dalit women in Kerala as a whole (data for Malabar alone are not available for Dalits), the literacy rate was only 17 per cent in 1961. It was after the implementation of land reform and a sharp rise in public investment in school education that literacy rates expanded among Dalits, especially Dalit women. The same holds true for Adivasi women also.
(Click to enlarge)

Take the case of IMR. In the early-1950s, the IMR in Kerala was 120, while that in India was 140; in other words, they were largely comparable. Similarly, as an average for 1921-30, the life expectancy of birth in Kerala (the composite state) was 30 years for men and 33 years for women. The corresponding figures for India as a whole were 27 and 26 years respectively. Do these figures indicate that Kerala was far ahead of India in the social sector at the time of independence itself? Not at all. While specific inspiring periods of progressive social policy in the 19th and early-20th century cannot be underestimated, Kerala leaped ahead of other Indian States only after 1957.

These facts cannot be hidden by simply stating that Kerala “started at the highest level at independence”, as Panagariya does.

Thirdly, Panagariya rejects the role of “successful public sector interventions in education and health as the source of sustained high levels of education and health in Kerala”. One, about 53 per cent of children in Kerala between ages 7 and 16 study in “private schools”. Two, he says that public expenditure on health in Kerala is barely 1 per cent of its GSDP. Three, private expenditure on health care far exceeds public expenditure on health care. From these three points, Panagaraiya argues that “the conventional and dominant story of Kerala as a state-led success crumbles in the face of hard facts.” One only wishes that Panagariya was less economical with facts and figures in his argument.

Let us take education first. Public expenditure on education has risen consistently in real terms in Kerala over the last six decades. As long ago as 1960-61, total government expenditure on education in Kerala was 3.7 per cent of GSDP. Starting at close to 4 per cent of GSDP in the early 1960s, public expenditure on education rose to a peak of 6.5 per cent in 1986-87, and has fluctuated between 5.5 and 6.5 per cent since then, along a marginally declining trend (recent data on public expenditure are biased downwards by the fact that there is no reliable data on how much the panchayats spend on education).

Most remarkably, about 96 per cent of all schools in Kerala are funded by the state. Panagariya’s generalised argument about “private schools” in Kerala arises from a lack of knowledge of the way school system is organised in the State. The Government of Kerala funds two types of schools. The first are schools established, owned and run solely by the State government. Together, they constitute around 36 per cent of all schools in Kerala. The second type is “aided” schools, which are owned and managed by private agencies. Here, the government meets the major component of their annual expenditure, namely, salaries. Aided schools also receive grants-in-aid from the State government for buildings and establishment, teaching and instructional material (including libraries and laboratories), and recreational facilities. This category covers 60 per cent of schools in the State and predominates at all levels of schooling. The last category, fully private schools, covers only about 4 per cent of all schools in the State.

It should then not be surprising that a large share of children study in the government-aided schools. Panagariya’s attempt is to portray government-aided schools also as private schools, and try to prove that the government does not have any role in their functioning. That is, at best, disingenuous.

Indeed, aided schools have been an integral part of the path in which Kerala’s educational system has evolved over a period. It is a legacy of the important role that social and religious movements played in Kerala’s educational history. Before 1957, a large share of these schools were owned and managed by educationally privileged communities like Nairs and Christians; thus, socially backward communities were largely under-represented in the educational system.

The educational policy of the Communist government in 1957 tried to change the educational system in two ways: one, by expanding the public school system by opening new government schools; and two, by trying to socially regulate the activities of private schools. While there was much success on the first count, the second has had a roller-coaster ride given the strong resistance put up by the privileged communities and their organisations. The present classification of schools into government/aided/private is an outcome, however imbalanced, of this roller-coaster political ride over more than 50 years. Even today, there is much in the aided school sector that can be termed “exploitative”, and there has to be additional social regulation of their activities. Yet, the larger point to note is that the partially-successful effort to socially regulate private schools in Kerala has ameliorated much of the damaging consequences of unfettered privatisation of schools.

Let us now take the health sector. There is no question that the foundations of Kerala’s phenomenal achievements in health were laid by its wide network of public health system. Panagariya chides Kerala’s public expenditure on health at 1 per cent of GSDP, even while it was the highest for any State in India. Even as he does not use any benchmark to judge Kerala expenditure standards, he ignores one of the important points argued out by Amartya Sen on whether public spending on education and health can be “afforded” by poor countries:

The viability of this “support-led” process is dependent on the fact that the relevant social services (such as health care and basic education) are very labour intensive, and thus are relatively inexpensive in poor - and low-wage - economies. A poor economy may have less money to spend on health care and education, but it also needs less money to spend to provide the same services, which would cost much more in the richer countries. Relative prices and costs are important parameters in determining what a country can afford.”

Even as its public expenditure was 1 per cent of its GSDP, Kerala was able to significantly expand its public health network by the 1970s itself. The commitment of public expenditures helped Kerala to expand health services and facilities equally among the rural and urban areas. Thus, by the late-1980s itself, about 70 percent of all hospitals and dispensaries as well as about 52 per cent of all hospital beds were located in rural areas. Kerala was the only State where the share of hospital beds in the rural areas was above 50 per cent. Such a phenomenal expansion of health care and services into rural areas would have been unthinkable without adequate public expenditure. Panagariya’s article is ignorant or dismissive about this basic fact.

The higher share of population in Kerala choosing private hospitals over public hospitals is a more recent trend and is indeed worrisome. However, even here, the argument is hopelessly out of context in Panagariya’s article.

It need not be surprising that in any society where public health has historically focussed more on basic/primary health services, more people will chose the private sector for secondary and tertiary health care. Kerala’s example is no different. Thus, in the initial phases of public health expansion in the State, there was no significant difference between the growth of private and public expenditures. Between 1961-62 and 1973-74, while per capita health expenditure of the government grew by 1.4 per cent, per capita health expenditure of the private sector grew by 1.9 per cent. However, it was after the mid-1970s that the differential begins to rise sharply. Between 1974-75 and 1986-87, while per capita health expenditure of the government grew by 2.5 per cent, per capita health expenditure of the private sector grew by 4.9 per cent. This differential has further increased in the more recent years, which is what has excited Panagariya.

Why has this happened? One, beginning from the mid-1970s, Kerala witnessed a sharp rise in disposable incomes, thanks to the remittances from the rising share of workers migrating to the middle-east. This put more money, on an average, in the hands of people who chose private hospitals over public hospitals. Two, the sharp fiscal crisis of the state from the 1980s onwards pre-empted any further expansion of public investment into the secondary and tertiary sectors. Even at the primary level, conditions worsened in the absence of fresh investment. With more disposable incomes and no public hospitals to meet the rising demands, the private sector came in quickly to fill the gap. That is why private health expenditures exceed public health expenditures, as a share of GSDP, in Kerala today.

In fact, what Panagariya posits as a positive feature – the expansion of private health care – is what is precisely wrong with Kerala’s health care system today. Every study on the health sector in Kerala would tell us that out-of-pocket health spending of people in Kerala has risen sharply in the last two decades. This demands a quick expansion of public expenditure into the secondary and tertiary sectors of health care. What constrains the state’s capacity to do so is a direct fallout of the post-1991 economic reform process, which Panagariya and his like enthusiastically support: fiscal austerity. The insistence on fiscal austerity at the central level has shrunk the finances available with States. Thus, funds allocated for the payment of salaries of health care staff and purchase of drugs have been forced to be cut, leading to a deterioration of the overall quality of public health services. It is funny now to see the same neo-liberals, who encouraged a cut in public expenditures, turn around and blame the public sector for poor quality services!

It may be fashionable for neo-liberal economists to run down, or make fun of, the developmental achievements of Kerala or the role of the Left in that long-drawn process. Of course, any critical view of Kerala’s experience should be welcomed. However, Panagariya’s so-called “critical” view is blinded both by dogmas of growth fetishism as well as erroneous use of statistics. Sorry, not welcome!

R. RAMAKUMAR


RAmkumar's blog

1 comment:

  1. This is with reference to the article “Cracking the Kerala Myth” by Arvind Panagariya (Professor of Economics, Columbia University), that appeared in The Times of India, dated 2nd January 2012. It may eminently suit the political convictions of the Columbia Professor to criticise Kerala’s development experience from his free-market perspective; he is also free to argue that Kerala’s is not a “state-led success” and that its “left-of-centre governments” did not contribute to its success. However, while doing so, one expects a modicum of rigour in argument. Sadly, Panagariya’s piece falls flat, both in its historical grasp and statistical rigour.

    ReplyDelete