Punditry has it that what gets measured gets managed. In the post pandemic world though measurement is massaged and managed. Perceptions of political success apparently depends on comparative failures, doing well depends on others doing badly. Measures therefore are twisted, tortured and tabulated to testify triumph. Data is the new oil. It is also the new snake oil.
This week, US President Donald Trump illustrated the phenomenon with an epic tweet. “For the 1/100th time, the reason we show so many Cases, compared to other countries that haven’t done nearly as well as we have, is that our TESTING is much bigger and better. We have tested 40,000,000 people. If we did 20,000,000 instead, Cases would be half, etc. NOT REPORTED!”
The reality is that tests only throw up what is there…that is the level of transmission and infection present at the given time in any geography. Tests are not the problem, they are a part of the solutions. The tweet is reminiscent of ‘Ravens Paradox’, a paradox presented by Carl Gustav Hempel in the 1940s on what constitutes as evidence for a given conclusion.
The saga of false positives can be traced back to the onset of the pandemic, in China’s denials and deception on the Wuhan Virus followed by the dereliction of duty of the WHO which said it found “no clear evidence of human-to-human transmission of the novel #coronavirus” and delayed till March 11 the decision to declare a global pandemic.
The quarantine of logic and counterfactuals is not unique to the United States. India has now crossed the 800,000-mark and is heading for the million milestone. Politicos who normally stick to absolute and gross numbers have figured out that normalising case count to population or case per million alters the optics. The United States which tops the table has 9,100 cases per million persons, Brazil which is on second spot has just over 8,000 per million and Russia which is fourth on the table has 4,900 cases per million. India, in contrast, has just 575 cases per million.
What is left unsaid in the comparison is the low level of testing or the ratio of tests per million persons. Testing in the United States is around 124,000 per million, in Brazil it is over 21.150 per million. In India, testing currently hovers at 8,000 per million — that is, 11.02 million tests in a country with over 1,350 million people.
The picture worsens at the state level. As per the Niti Aayog data, testing in ten large states is lower than the national average — Bihar which saw the return of 25 million migrant workers has conducted 2,100 tests per million, Telangana is at 2,637 per million and Uttar Pradesh around 3,800 tests per million. Many state governments harboured the notion of ‘lower testing means lower infection’, curtailed testing and even declared victory triggering rah-rah rhetoric.
The misplaced notion though is subject to the statute of limitations. Reality dawns with the lifting of the lockdown. While Maharashtra, Tamil Nadu and Delhi lead the case count and account for over half the cases, the true extent of transmission in states such as Gujarat, Telangana and West Bengal is showing up even with incremental rise in testing.
Politicians have expanded their tool kit and frequently present a dip as a trend. On June 18, West Bengal declared 515 recoveries and 391 new cases. CM Mamata Banerjee cited this as a positive. On July 10, the state reported 1,198 new cases. The success in containing infections in Mumbai’s Dharavi deserves plaudits but it was generalised. On Friday, Maharashtra announced the third lockdown of industrial hubs Thane, Pune and Pimpri-Chinchwad as it recorded over 7,800 new cases. This is not unique to Indian politicos, many elsewhere have declared triumph only to find the facts catching up.
A factoid frequently cited as a positive is the ratio of recovered persons to total infected – nearly 5 lakh of the 8 lakh infected have recovered. It is true that India has a better rate of recovery among those infected and this is partly explained by the younger demography. It is also true that the death toll has been low at barely 16 per million. Globally, fatality rates fluctuate with variables — particularly with age of population, on what is called comorbidity or pre-existing conditions of those infected.
Conditions for comparisons though must be comparable — that is registration of deaths and the cause. Record keeping in India though is a known unknown. A 2019 Home Ministry report has observed that analysis of state-level data reveals that 21.9 per cent or one of five deaths were simply not registered. It gets worse. A report of the Registrar General of India on reporting of deaths places the percentage of medically certified deaths to total registered deaths at 22 per cent — effectively the cause of death in nearly eight of ten cases is not known or at least not registered.
The virus has persistently defied definition and this has triggered hypotheses on failures and of successes. To paraphrase the Bard in Julius Caesar, politicians have and will peddle snake oil, in the hope that people will willingly believe what they wish to be true.