A nationwide blackout in India recently left some 600 million people without electricity. Given that a large number of oDesk contractors are from India—it is the second-most active country by hours worked—oDesk Staff Economist John Horton set out to see if the effects of the blackout could be seen in our data.
Here’s what John did:
- He checked the number of hours worked and number of job applications submitted by contractors in India for each day during the past month.
- He divided these two daily counts (hours worked and jobs submitted) by the respective totals of each for all work being done on oDesk worldwide.
The results give us a sense of what was supposed to happen on a given day, so we can compare what was supposed to happen to what actually happened during the blackout.
Some of John’s resulting observations:
- There is a notable drop-off in the hours worked. John’s eyeball calculation indicated that contractors in India should have been responsible for about 22% of the hours worked on the day of the blackout, while the actual number is closer to 17.5%. This is less of a fall-off than might be expected given the “Half of India Without Power" headlines. This may be explained by access to private generators, or perhaps oDesk is overrepresented in parts of the country that were less affected by the blackout.
- There is no corresponding obvious drop-off in the fraction of job applications. Perhaps unaffected contractors made up the difference and took advantage of the now-thinner market. If anyone has data on what parts of the country were affected by the blackout, we could test this notion.
- Contractors in India take weekends off, both in terms of working and applying to jobs (or at least more so than their oDesk counterparts from other countries). However, this time series shows the fraction for a given day, so there's no reason for a strong weekend/weekday pattern. See oDesk Country Explorer for more of this kind of data.
- Indian contractors are generally overrepresented in the application pool, making up approximately 25% of applications but only about 20% of hours worked. However, this could easily reflect differences in the kinds of categories Indian contractors work in—there is a great deal of variation in the average number of applications per opening across the different job categories.
John’s Results Visualized:
The time series for job applications (top) and hours worked (bottom) are plotted below [see here for the code for the plots], with the second day of the blackout annotated in red.