I read an interesting data analysis by Jonathan Regenstein last week. It was an analysis of unemployment insurance claims in Georgia. The data visualization vividly shows how bad the labor market is during the COVID-19 Pandemic. https://rviews.rstudio.com/2020/04/16/state-unemployment-claims/ After reading Georgia’s case, I got curious about my own state, New Hampshire.
I created the plot below by borrowing the code from Jonathan Regenstein’s post. The plot displays the highest weekly employment insurance claims every year since 1999. The data was retrieved from the Saint Louise Federal Reserve Bank database (FRED).
- The big spike in April of 2020 shows just how bad the unemployment siutation is in New Hampshire during the Coronavirus Pandemic. It dwarfs the highest unemployment claims week in the past two decades, including the Great Recession of 2007-2009.
- Prior to the Pandemic, the worst week every year has been in either December or January, with an exception of 2015 when the highest unemployment insurance claims were in February.
We just saw that New Hampshire was hit hard by the COVID-19 Pandemic. How about other states? How does New Hampshire compare against other states? The plot below shows the number of unemployment insurance claims in New Enlgand in 2020. All six New England states have huge spikes in claims starting on March 21, which I believe coincides with the beginning of the lockdown order.
- Massachusetts dwarfs other states in New England. This is expected because of its relative size to other states in the region.
- The claims peaked at the end of March in Massachusetts, while the peak doesn’t arrive until after April in three northern New England states. It is likely indicative of the important role that Massachusettes plays in the region’s economy. For example, New Hampshire’s tourism industry heavily relies on tourists from the Boston area. The industry is a major employer in the state, particularly its northern regions.
However, we should account for each state’s population to make a fair comparision among states. One solution is to compare percent changes over time, instead of number of claims. The interactive map below displays percent changes during the lockdown from the same time in the prior year. More specifically, it aggregates weekly unemployment claims for the period of second half of March 2020 and first half of April 2020, aggregates claims for the same period in 2019, and calculates the percent changes between the two periods.
- New Hampshire recorded more than 125,000 claims since 03/21/2020, which is an increase of 6,400 percent from 1,927 claims during the same time in 2019. This is the largest increase in the country by far during this period. The next highest was an increase of 5,000 percent for Louisiana.
- Although it’s not clear why this is the case without further investigation, it may have to do with the industry composition of New Hampshire’s economy, which relies on industries that don’t allow for working from home (e.g., tourism and manufacturing). In addition, it likely reflects how well the state was doing prior to the Pandemic.
We can also control for the size of states by dividing the number of claims by state population. The interactive map below displays the number of claims per 1,000 working-age adults since March 21, 2020 when the lockdown started having effect on the labor market. I grabbed the population data from the United States Census, 2014-2018 American Community Survey.
- New Hampshire has 114 claims per 1,000 people of 16 years or over between March 21, 2020, and April 11, 2020, which is the fifth highest in the country.
- Nevada has the highest claims per 1,000 people of 16 years or over during the same period, with 132 claims.
- As of April 11, 2020, the Southwest and Rocky Mountains Regions seem to be doing better than the rest of the country on a per-capita basis.