Body
Most countries began implementing stringent measures to reduce the spread of the COVID-19 virus in March, including school and workplace closures, limits on public gatherings, and travel curbs. But did these measures actually achieve the result of restricting people moving out and about in the face of the disease? Thanks to smartphone data from Google, it is possible to measure the impact of pandemic-related lockdowns on personal mobility.
The data presented here indicate that people's compliance with lockdown orders varied across income groups. Individual mobility decreased more in higher-income countries than in lower-income countries with similarly stringent measures. The greater the shares of extreme poverty and vulnerable employment,[1] the more likely people were compelled to move about out of economic necessity. The factors in mobility were also driven by demographics, and the number of new cases and new deaths per million inhabitants had an impact on mobility.[2]
How lockdown policies and individual mobility were measured
To analyze these phenomena, this blog post uses the Oxford COVID-19 Government Response Tracker stringency index to compare lockdown policies. This index considers nine aspects of stringency: school closures, workplace closures, cancellation of public events, restrictions on gatherings, public transportation closures, stay-at-home requirements, restrictions on internal movement, international travel controls, and public information campaigns. The more stringent the measures are, the higher the stringency index.
Individual mobility is measured by Google (smartphone data). The GPS data Google collects from users who have opted in to storing their location history on their phone's Google account paints a picture of individual movement during the pandemic. Individual mobility is defined in this post as an average of movement associated with four types of locations: retail and recreation, groceries and pharmacies, transit stations, and workplaces. The values of these indexes are expressed in comparison to a baseline value that corresponds to the median mobility in January 2020 for each country.
Data reveal stringency measures reduced individual mobility but with significant variations across income groups
First, data on 118 countries[3] reveals that all countries implemented stringency measures in March 2020, with some differences across income levels in countries. On average between March 25 and May 9, middle income countries implemented the most stringent measures followed by high income countries and low-income countries (figure 1, panel a). Meanwhile, individual mobility decreased most significantly in upper middle–income countries, followed by high-income countries, lower middle–income countries, and the least in low-income countries (figure 1, panel b).
By comparing the two variables, it appears that stringency measures reduced individual mobility but with significant variations across income groups. People in higher-income countries responded more strongly to stringency measures than people in lower-income countries did. For instance, the stringency index for Belgium (high-income country) was equal to 79 and individual mobility decreased by 58 percent on April 6, while in Kenya (lower middle–income country) on the same day, the stringency index was equal to 85 and individual mobility decreased by only 33 percent.


Factors that help explain why individual mobility decreased less in the less wealthy countries
What explains these different responses? The answer lies partly in the perceptions of disease risk as measured by the number of new cases and new deaths per million inhabitants per day. Hale, Angrist et al. (2020) point out that countries implemented more or less similarly stringent policies in mid-March regardless of the timing of the domestic caseload. The number of COVID-19 cases and deaths occurred at first mainly in higher-income countries (figure 2). The fact that the virus was less widespread in lower-income countries than in high-income countries explains in part why individual mobility decreased less in the less wealthy countries despite similarly stringent measures.
Other factors include the share of extreme poverty, the share of vulnerable employment, the share of employment in agriculture, and the share of the population that belongs to a demographic known to be more vulnerable to the coronavirus (those aged 65 or older). These factors prove important in explaining why the decrease in individual mobility was different across countries.
Table 1 gives a description of these factors by income group. Unsurprisingly, the shares of extreme poverty, vulnerable employment, and employment in agriculture are much higher in lower-income countries than higher-income countries. Stringency measures are less effective in countries with high shares of people living in poverty or who have vulnerable jobs; both groups must continue working—i.e., stay mobile—to avoid dire living conditions as they are not usually protected by social safety nets.
In addition, the majority of people in low-income countries—55.36 percent—are employed in agriculture, compared to 3.55 percent in high-income countries. An increase in the share of employment in agriculture is associated with an increase in mobility, since activity cannot be reduced in this sector. The share of vulnerable employment and the share of agriculture are highly correlated with a correlation coefficient of 0.90.
|
High-income countries |
Upper middle–income countries |
Lower middle–income countries |
Low-income countries |
Extreme poverty |
0.68% |
3.32% |
14.84% |
45.3% |
Vulnerable employment |
10.43% |
27.84% |
55.84% |
78.34% |
Share of population aged 65 or older |
15.16% |
8.77% |
4.67% |
3.05% |
Share of employment in agriculture |
3.55% |
17.03% |
37.13% |
55.36% |
Meanwhile, the share of the population aged 65 or older is five times higher in high-income countries than in low-income countries. Thus, demographics also play a significant role in individual mobility: A higher share of this vulnerable population is associated with a higher decrease in mobility.
In conclusion, stringent measures adopted by countries have decreased individual mobility overall. But the decrease in individual mobility in response to lockdowns with similar levels of stringency has varied by income group. The response has been weaker in lower-income countries. This can be explained through factors such as extreme poverty, perception of risk, share of vulnerable employment, and demographics. Policymakers need to take these factors into account to effectively implement lockdown measures, regardless of their level of stringency.
Notes
1. The International Labour Organization defines vulnerable employment as “the sum of own-account workers and contributing family workers. They are less likely to have formal work arrangements, and are therefore more likely to lack decent working conditions, adequate social security and ‘voice’ through effective representation by trade unions and similar organizations.”
2. These results are confirmed by using panel regressions (see forthcoming Policy Brief on the same topic).
3. These include 47 high income countries, 33 upper middle–income countries, 27 lower middle–income countries, and 11 low-income countries.
Author's note: I thank Olivier Blanchard, Pinelopi K. Goldberg, and Adam Posen for their invaluable guidance and support.