Getting reliable and timely data for unemployment in India has been a difficult proposition till now, since the government releases data with a considerable lag and there has been precious little non-government effort towards data generation. Now, Centre for Monitoring the Indian Economy (CMIE) has begun tracking unemployment in India since January this year, making the study on job creation in India a little easier.
Mahesh Vyas, CEO & MD, tells Sindhu Bhattacharya, in an e-mail interview that the unemployment rate did not fall right after demonetisation announcement on 8 November. In fact, unemployment rate had been dropping for some time. It fell from 7.4 percent in the week of the announcement to 6.2 percent, and in the week of 27 November, it dropped to less than 5 percent. But it has risen thereafter - to 6.1 percent in the week of 4 December, to 6.6 percent in the week ended 11 December and then to 7 percent in the week ended 18 December. Execrpts:
When did CMIE begin tracking unemployment in India? What spurred you to start this survey?
We started tracking unemployment in January 2016. The series from January, was released in April 2016. Since then, we have been releasing the unemployment rate every day.
Unemployment is an important lead indicator of any economy. In fact, inflation and unemployment are the two most important fast frequency measures of a modern economy. India did not have a fast-frequency measure of unemployment. In fact, India's unemployment numbers were very outdated and of practically no use for policy making or for business. Data gaps impair our understanding and they lead to unsubstantiated claims.
The availability of reliable data saves us from avoidable and often acrimonious debates. CMIE decided to fill in this gap and decided that the unemployment rate must be made for a public good. BSE became a partner in this endeavor and helped make the unemployment rate a public good by becoming a sponsor. The enlightened view of the BSE board and CMIE's large survey machinery has helped make India a much more informed economy today.
What is the methodology for tracking unemployment? How do you manage to get a daily report?
Unemployment is calculated by us, by using the standard formula used internationally, which is the ratio of unemployed persons to the total labour force.
Is the sample size big enough for a proper assessment of the jobs scenario? On what basis are interviewees selected for this data?
The sample used by us is the largest deployed in the country. It consists of nearly 160,000 households. The sample size for the last round conducted during May-August 2016 was 159,624. This provided the employment / unemployment status of about 580,000 individuals who are of 15 years of age or more. This is much larger than the sample size deployed by the National Sample Survey Organisation.
Such a large sample size enables us to get (a) a robust estimate of unemployment and (b) an in depth understanding of unemployment in India. Further, the field execution of the survey is organised such that we get a reasonably large sample that is well spread over rural and urban areas every day to make fast-frequency estimations such as weekly unemployment rate and a 30-day moving average daily estimate as well. The weekly estimation is based on a sample size of about 10,000 household or about 36,000 individuals of 15 years or more.
For the purpose of calculating the unemployment rate we seek information on all persons in the sample households who are of 15 years of age or more. The respondent is the most articulate person of the sample household. Such a person speaks on behalf (and often in consultation with) of all members of the household. So, the answer to your question is that all persons who are more than 14 years of age in our sample households are our interviewees. But, we do not have to interview all of them. One articulate member of the household answers on behalf of all interviewees.
Does the data provide a breakup of unemplyment in organised and unorganised sector? Why only rural vs urban?
The unemployment rate is computed based on a household survey. Households are not organised or unorganised. And, a household survey does not yield the unemployment rate in the organised sector or the unorganised sector. To get this, we need to conduct a survey of employment in enterprises from the organised and unorganised sectors.
However, households do belong to rural or urban regions. And so, it is possible to generate rural and urban unemployment rates. It is also possible to generate state-level estimates of unemployment , and we do produce these every month. At the end of every Wave (ie once in 4 months when the entire sample is surveyed), we also generate unemployment rates by age, gender, and by level of education. All these are estimated at the rural and urban levels of each state as well. A detailed statistical profile is generated every four months that provides much greater insights into the unemployment status in the country.
What has been the trend in unemployment this calendar year? How has it changed, if at all, after demonetisation?
The unemployment rate averaged 8-9 percent from January through September 2016. It fell to 6 percent during October and November.
The unemployment rate did not see any fall after demonetisation. This surprised us because there were several reports every day of broken supply chains and much stress among producers of goods and in the markets. However, the data did not show any increase in unemployment. Unemployment, in fact, dropped for some time. It fell from 7.4 percent in the week of the announcement to 6.2 percent and in the week of 27 November, it dropped to less than 5 percent. Unemployment has risen thereafter. It rose to 6.1 percent in the week of 4 December, to 6.6 percent in the week ended 11 December and then to 7 percent in the week ended 18 December.
Our experience in being able to track the impact of demonetisation on unemployment vindicates our view that a fast-frequency measure is critically important in this world where uncertainties seem to increase every day.
Will the scope of the survey be increased going forward?
Yes, it will be increased. This is only the beginning.
Does it take into account employment generated through welfare schemes like MNREGA?
Yes. It does take into account employment generate through welfare schemes. It does so by the construct of the questions. Welfare schemes are not specific questions. They are subsumed, though, in the questions asked.
Does it also capture movement in wages for organised and unorganised sectors?
It does not capture wages; it does not capture organised and unorganised sectors as clarified earlier.
Updated Date: Dec 21, 2016 13:57:56 IST