Billy H.C. Kwok/Bloomberg

OPINION – Measurement matters

It is not uncommon that the effects of economic contractions take some time to be visible. That happens both with measures of economic activity, such as the Gross Domestic Product, and with other related indicators such as, for example, employment levels or retail sales. After all, statistics take time to collect and process. 

After the pounding in the first half of last year, the economy recovered a bit but is roughly about half the size it used to be. The current economic trough has been going on for more than a year, and its effects should be visible in many other indicators. But, surprisingly, that does not seem to be the case in some unexpected instances.

Figures for overall employment and unemployment seem relatively immune to the ravage one would expect from a contraction of such magnitude. As more detailed data becomes available, some measure of bewilderment is not going away. Let us take a couple of examples from surveys made twice a year on some select economic sectors. 

In the case of casinos, surveys take place in June and December. Last June, the employment losses reported amounted to less than 1,500 workers, less than four percent of the labor force. The drop for full-time employees was marginally smaller, affecting just a tiny fraction of the entire staff. Moreover, according to the published figures, changes relating to working statuses other than full-time were barely significant. Yet, for the first eight months of this year, this is an activity that operated at less than one-third the level it reached before the pandemic. 

Also, could we expect retail trade, which boomed with the influx of visitors, to lose less than 2,5 percent of its full-time staff? Indeed, the sector relies on both internal and external demand. According to GDP data, the former has recovered most of the lost ground, but it is still some ten percent below the 2019 figures. And the number of visitors dropped precipitously. The first half of this year saw less than a fifth of them, in the same period.

Similar observations could be made about other critical or sensitive sectors. With different levels of measurable impact, at least as seen from the perspective of these surveys, the results are still surprising in one way or the other. We still must get plausible and testable hypotheses to determine if and why it is so.