Understanding CMIE's CPHS Employment Data

On February 27, 2021, the Economic Times carried an op-ed claiming that data from the Periodic Labour Force Survey (PLFS) better fit the empirical predictions of a macroeconomic theory called “Okun’s Law” than data from Consumer Pyramids Household Survey (CPHS). The authors concluded that since PLFS fit the empirical predictions while CPHS did not, the employment data from CPHS was not trustworthy. We wrote a full response to that piece, which was carried in the Economic Times on June 8, 2021 entitled “A Data Smell Test That Smells Fishy”.

In it, we pointed out five specific faults in their piece. Three of these were misrepresentations of the data and two were misunderstandings of the economic theory that was deployed by the authors. The same authors have posted a rebuttal to our response, carried in the Economic Times on June 16, 2021. However, they have not responded to our earlier criticisms. They make additional misrepresentations in their latest response, which we address here.

Misrepresenting CPHS

The authors write that our rebuttal does not address serious lacunae in the CPHS methodology, but fail to point to such a lacuna in the methodology we employ. This is because the CPHS methodology is well considered, carefully documented and trusted by academic researchers across the world. Rather, the authors continue to make factual mistakes about CPHS and the estimates that CMIE produces from it.

In their most recent response, the authors also claim that the CMIE employs inconsistent reference periods for different indicators about labour force participation. This is false and reflects their poor understanding of the methodology we employ.

In our last response, we pointed out that the CMIE employment rate cited by them did not match with CMIE’s published numbers. The authors responded to this as follows: “CPHS does not report the most important indicator of the labour market: employment rate”. This is extraordinary.

CMIE has been publishing the employment rate (ER) along with the labour participation rate (LPR) and the unemployment rate (UER) regularly.

Misrepresenting Okun’s Law

The authors relate both CPHS data and PLFS data to “Okun’s Law”. In doing so, they have made three errors in interpretation.

  1. “Okun’s Law” doesn’t talk about employment rate: “Okun’s Law” does not talk about the relationship between the employment rate and GDP as the authors say. Instead, it talks about the relationship of the unemployment rate with GDP. The authors claim that their findings hold even when using the unemployment rate although they provide no evidence.
  2. “Okun’s Law” doesn’t speak about nominal GDP: “Okun’s Law” is about change in real GDP and not nominal GDP but the authors used nominal GDP. The authors claim that their findings hold true even when using real GDP although they provide no evidence.
  3. Urban unemployment is not a proxy for overall unemployment: In their original piece, the authors tried to use measures of urban employment to proxy overall employment. They continue to do so, without meaningful justification.

This piece was corrected on June 23, 2021.