Number Theory: What does mobility data tell us about the state of the economy?
The Indian economy is “on the verge of normality”, if the latest reading of Nomura India Business Resumption Index (NIBRI) were to be taken as an indicator. NIBRI reached 98.1 (provisional value) in the week ending February 14, where 100 signifies pre-pandemic level. It was at 95.9 the week before that.
“Since its trough in April last year, NIBRI remained on an uptrend through 2020 that continued into Q1 2021 (NIBRI average: 94.7 from 86.5 in Q4 2020), reflecting fewer new pandemic cases”, said a note from Nomura economists Sonal Varma and Aurodeep Nandi. The note infers improvement in NIBRI as a sign of growing economic momentum, and predicts a 1.5% GDP growth in the quarter ending December 2020 and 2.1% in the March 2021 quarter. To be sure, the latest NIBRI numbers also indicate a moderation in labour participation rate (LPR), which is defined as the share of working age population working of looking for a job. LPR moderated 40.5% in the week ending February 14, compared to 40.9% the week before that.
An HT analysis of Google’s community mobility data can complement the economic normalisation story being offered by NIBRI.
Google Mobility data has been relatively flat since December 2020
While mobility has been increasing continuously, its recovery is relatively slower when compared to NIBRI. To be sure, mobility levels had reached 90% of pre-lockdown values (median value for the day of the week in the January 3, 2020 to February 6, 2020 period) in December 2020 itself. Average monthly value of all movement increased by a large magnitude between September 2020 to December 2020. This trend reversed itself to show a small decline in January 2021 and there has been a smaller increase in the February 2021 (data up to February 9) value. This trend remains the same if mobility on account of residential movement and visits to grocery stores and pharmacies (essential activities) are excluded. The government announced significant relaxations from lockdown related restrictions beginning February 1.
They included allowing states to determine the size of gatherings, which earlier had a cap of 50% of hall capacity or 200 persons in closed spaces. Cinema halls were allowed to operate at 100% capacity compared to the 50% cap that was imposed when they were allowed to open from October 15. All kinds of exhibition halls, apart from Business to Business exhibition halls that were allowed to open earlier, were allowed to function. Swimming pools were similarly permitted for use of all compared to being restricted to sportspersons earlier. This suggests that the existing marginal deficit in mobility could be a result of self-imposed restrictions rather than official curbs.
Restoration of mobility differs across activities
The overall mobility index used here – which measures visits to and length of stay at particular place as a percentage of pre-lockdown levels -- is an average of mobility for visits to six kinds of places: retail and recreation, grocery and pharmacy, parks, transit stations, work and residential. Each of these activities shows different levels of normalisation. Latest retail and recreation mobility index shows the lowest recovery, with activity levels at just 75.2% of the base value. Work related mobility was at 83.9% compared to 90.2% for transit. The relatively muted recovery in work related mobility is in keeping with the weakness in labour markets captured in other high frequency indicators as well. For example, both the manufacturing and services Purchasing Managers’ Index (PMI) noted a fall in employment in January, even though the pace of job shedding has been moderating.
Grocery and pharmacy head shows that visits to such places has more than recovered from the base value, with the latest value at 114.6%. Residential mobility continues to go down, suggesting that people are spending less time indoors, although it is still higher than the base value.
Work-related movement low in most rich states
Visits to workplaces are farthest from pre-pandemic levels in some of the richest states -- such as Delhi, Punjab, Haryana, Gujarat and Maharashtra – and the closest to pre-pandemic levels in some of the poorest state, such as Bihar, Andhra Pradesh, Uttar Pradesh, and Assam. However, there are exceptions. Work-related mobility in Chhattisgarh was 82.9% of the baseline, only higher than that in Punjab (81.1) and Delhi (75.9%), among major states. On the other hand, in Karnataka, the mobility levels in February are behind only the four states with the highest mobility levels listed above.
Mobility levels not dependent on caseload
Another interesting aspect of the mobility levels across states is that they do not appear to depend on how the state is faring in managing Covid-19 infections. Active cases per million population in Delhi were 63 on February 9, 6th highest among eighteen major states. But non-essential movement in Delhi in February (until February 9) is the lowest (78.9% of pre-pandemic levels). On the other hand, Kerala had the highest number of active cases per million population on February 9 (1927), but only the eighth lowest mobility level (90.3% of baseline) among major states in February.