Essay: How Coronavirus Impacted Labor Demand and Markets: An Analysis of Burning Glass Technologies Vacancy Data and UI Claims

3
1. Introduction
“By labor is meant the economic work of man, whether with hand or head.” – Prof. Marshall
Labor is one of the four factors of production without which the process of production is impracticable. Labor markets, markets for the employees and employers, are influenced by the forces of supply and demand. Labor demand is a concept that defines the amount of labor that an economy is willing to employ at a certain time. Economic changes have an important effect on the Labor market. Rapid economic growth will create a plethora of new employment opportunities. Correspondingly, an economic recession can have a severely detrimental impact on labor.
In December 2019, coronavirus disease, a highly contagious disease was identified. It spread rapidly all over the world in the span of eight months. The outbreak of the Coronavirus resulted in a significant contraction of economic activity. This is due to people altering their actions in order to adapt to the risky circumstances and the constraints imposed by governments.
The literature review focuses on how coronavirus affected the demand for labor and the overall labor market. It begins with the analysis of the United States job vacancy data obtained by the Burning Glass Technologies (BGT), the initial claims of Unemployment Insurance (UI) and the Bureau of Labor Statistics (BLS) to examine the impact of coronavirus on the labor market. The variations in the impact of the pandemic will be studied. Then, the labor market shocks that individuals experience will be evaluated followed by an assessment of the steps they adopted to cope with said shocks.
4
2. Data Analysis
E. Forsythe et al. (2020) in their study, to provide a more comprehensive account of how the labor market deteriorated over March and April 2020 in the US, evaluated both UI claims data and vacancy data from the BGT and confirmed their analyses with the help of the BLS employment situation data.
A. Burning Glass Technologies vacancy data
BGT is an information software firm that offers real-time data on employment analytics and labor market trends.
In the BGT vacancy data, a sharp drop in job postings commencing in mid-March was observed. The number of weekly postings declined by 44%, from 8,15,000 in the week of March 15 to 4,60,000 in the week of April 26.
B. Initial Unemployment Insurance claims data
UI claims data are beneficial in that they can reflect the universe of claims processed. They found that the drop in vacancies and employment coincide with a dramatic increase in the UI claims commencing in the week of March 15th. Claims peaked in the last week of March and deteriorated steadily over April. However, initial claims persist at historically high levels. More than 30 million initial claims were filed around mid-March and the second half of April.
Bartik et al. (2020) in their research paper measured the collapse and partial stabilisation of the US labor market from March to July. It recovered steadily from April through June before plateauing and declining again as the virus surged. The recovery, though very partial, allowed many employees to return to their former workplaces within a few months.
C. Bureau of Labor Statistics employment situation data
The BLS surveys supply the conventional and most closely tracked labor market well-being indicators.
5
It was observed that the rate of exit from employment to unemployment, which comprises of the workers that are unable to apply for benefits, rose from 3.5% in February to almost 20% in April. A 13% decline in employment was also observed.
Coibion et al. (2020) observed that the employment-to-population ratio has decreased significantly (≈8%). Their surveys indicate a 20 million decline in the employed workers. Interestingly, they found a less than proportional rise in unemployment, suggesting that most of these newly unemployed workers aren’t searching for new jobs.
3. Variations
Mealy et al. (2020) mapped the Remote Labor Index to industry categories to estimate industry-specific supply shocks from social distancing measures and quantitatively predict that: essential industries wouldn’t have any supply shock, health related occupations will experience increased demand but no corresponding increase in supply.
E. Forsythe et al. (2020) in their research observed variations in the impact of Covid-19 across different states and industries which are well in line with the predictions. Factors that cause these variations include Government policies, nature of industries with respect to elasticity of demand of the goods/services produced and work-from-home (WFH) measures facilitated by companies.
A. Across USA States
They identified that the timing and degree of the labor market responses had only some minor differences across states that varied in how late they implemented stringent social distancing measures. The states that introduced earliest measures appeared to experience the strongest impacts for each labor market metric, while those that never implemented controls saw milder impacts.
They then considered the relationship between the initial spread of coronavirus and the degree of the country-wide collapse in the labor market. They identified that states with
6
earlier spread of the pandemic have poorer labor market results. In sum, the decline of the labor market is driven by a national crisis. State-level policy heterogeneity is less significant than the spread of the virus itself coupled with rising uncertainty, disordered supply chains, and the decrease in demand for final products.
B. Across different industries
Job postings and employment both display sharp decreases in labor demand in April. The worst declines were in leisure and hospitality and non-essential retail. Job postings in both dropped below 50% of their February level. Parallelly, over March and April, postings in essential retail rose and employment remained largely flat. For nursing, the decline in postings was shallower and slower, but by April the number of ads in nursing fell by 25%. Postings and employment declined more in non-essential jobs than in essential jobs, as anticipated, with healthcare in between.
UI claims in essential retail and nursing increased the least, however, even their claims have risen by orders of magnitude, and these groups are a small fraction of the workforce. Thus, despite substantial differences across industries, UI claims indicate broad-based increases across both essential and non-essential industries, as well as in healthcare.
C. Work-from-home capability
The government restrictions make it difficult for non-essential employees who cannot WFH to keep up their work. Researchers reported that only about a third of employees have jobs where WFH is possible and found that it’s similar across essential and non-essential sectors, excluding health. UI claims rose by more for those who are unable to work from home. Initial claims increased about two-thirds as much for occupations that are WFH capable.
Intriguingly, postings tell a different story. Postings decreased by a similar amount in non-essential industries, regardless of whether WFH was facilitated. In essential sectors, the non WFH facilitating jobs saw a smaller decrease in postings. Although the Employment and Separations series indicate that employers eliminated more workers who are unable to WFH, vacancy postings decreased by a similar amount in WFH-enabled occupations. The decline in the labor market impacts all sectors, irrespective of their labor restrictions.
7
Although WFH-enabled occupations may be insulated from some of the labor demand shocks arising from stay-at-home orders, they still struggle from the negative aggregate demand shock stemming from the crisis.
4. Labor Market shocks experienced by individuals
Crossley et al. (2020) in their research showed that the magnitude of the economic effects of the COVID shock in the UK varies for different individuals. These variations arise partly because the direct impact differs depending on individual characteristics and nature of job, and partly because people take different steps to minimise the shock. The same economic shock has different effects.
The UK went into ‘‘lockdown” on 23rd March and the economy contracted significantly in March and April.
It was observed that the young and those without any assured work hours witnessed major declines in working hours at the beginning of COVID resulting in large falls in household earnings. By the end of May, the decrease in working hours for these groups had been partly reversed, and household earnings showed less cumulative decreases than for other groups. Sharp initial decreases in the fraction of working hours and in employment were seen for those aged under 30, but again has rebounded a little.
Black, Asian and Minority ethnic (BAME) groups witnessed different labor market shocks than non-BAME groups. The reduction in the fraction of working hours was similar between both the groups, but the reason that hours decreased varied significantly. While the decline for non-BAME groups was primarily driven by being moved onto the Job Retention Scheme (whereby workers were ‘‘furloughed” by their firm) rather than into unemployment, the decline among BAME groups was driven in equal measure by being moved into unemployment as they are 15 percentage points less likely to be supported by
8
the Scheme. BAME groups experienced a significantly larger decline in employment than others.
Those in the bottom three pre-pandemic average income quintiles witnessed the greatest declines in the fraction of working hours. Mealy et al. (2020) predicted the same.
Steps taken to cope with the shocks-
1. For many, impact was mitigated by the Job retention scheme
2. Mitigation of losses in earnings by the young was through private savings and by finding new work
3. For those in lowest quintiles, losses were mitigated by borrowing and transfers from family and friends
5. Findings
1. There is evidence that labor markets in states collapsed substantially due to the severity of coronavirus and not solely due to government restrictions.
2. Economic damage cannot be reversed simply by lifting stay-at-home orders.
3. Non-essential industries experienced broader retreat than essential industries.
4. Most of the newly unemployed workers aren’t looking for new employment opportunities
5. Young workers, BAME groups and bottom three average income quintiles witnessed major decline in working hours.
6. Conclusion
The unprecedented outbreak of Coronavirus resulted in a Labor market collapse. Both essential and non-essential industries were affected but the latter saw a more substantial
9
decline. Essential retail saw no decline in postings and small spikes in claims thereby being more likely to significant labor reallocation. WFH occupations were sheltered from immediate unemployment. The Job Retention Scheme failed to help the workers from BAME groups, working hours of young employees decreased and bottom three pre-pandemic average income quintiles were the worst affected. The employment levels plummeted, the demand for labor declined and people witnessed decrease in their working hours. The severity of the spread of coronavirus in States is directly proportional to the impact on their labor market. It was found that early implementation of stay-at-home orders had little to do with said collapse thereby ruling out the possibility of stabilising the economy just by lifting the restriction.