This post was co-authored by Daniel Sternberg, Sarah Gustafson, and John Schlerf.
Gusto enables more than 100,000 small businesses across the United States to take care of their teams with full-service payroll, benefits, compliance, expert HR, and more. This report analyzes leading indicators and changes to employment at small businesses for the full month of July 2020.
Net employment at small businesses continued to increase in July ‘20, but the recovery that started to take place in May and June has slowed down significantly. The percentage growth rate for total small business headcount was cut in half from June ‘20 and was two-third less than May ‘20. And hiring and rehiring of former employees has started to stall in July, likely as many small businesses take a wait-and-see approach to better understand the impact from the surges in COVID cases and what additional government aid will be passed with the original PPP expires. The positive change in net headcount in July was propped up by lower than usual terminations for this time of year.
While employers took fewer actions that negatively affect employees wages (furloughs, terminations, reductions in hours and wages), this may only be a temporary reprieve unless the government takes swift action to deliver more aid. Data from a study Gusto conducted in July shows that the clock is running out, with 50% of small businesses reporting that they have 6 months or less to stay in business.
The findings reported below represent data points with meaningful sample sizes[1], but based on the distribution of Gusto’s customers, they may not be fully representative of the industry and geographic breakdown of small businesses across the United States.
Quick Links
Key Findings
- Recovery growth rate has stalled, especially for the industries hit hardest by the pandemic: July ‘20 saw a headcount growth rate of 1.2%, compared to 2.5% growth in June and 3.4% growth in May. Industries hit the hardest by COVID have seen their growth rates conspicuously level off over the past month. Food & Beverage companies are still operating at 7% below their pre-COVID February headcount and 155% below expectations based on prior year data. This plateau is primarily driven by softened hiring, as terminations are now consistent with or slightly below previous years.
- Headcount remains significantly lower than what would have been expected without COVID: Overall small business headcount dropped precipitously starting in mid-March, and has been on a slow climb to that first week of March baseline in recent months. In July, total headcount was +2% above compared to pre-pandemic levels. However, when compared to normal headcount growth rate for spring and summer in recent years, July ‘20 expected was -78% below expected headcount.
- Less than half of furloughed employees have returned to work and often at less pay than pre-COVID: Only 37% of workers first furloughed in March ‘20 and 47% in April ‘20 have returned to work by July ‘20. Nearly 25% of those furloughed in March ‘20 that returned to work had their wages reduced by 10% or more compared to pre-COVID wages. Hourly workers were more than 4 times more likely than salaried workers to be furloughed and 3.5 times more likely to have had a reduction in hours of more than 10% in the month of July than salaried workers.
July ’20 Small Business Trends
Small business headcount grew for the third consecutive month, with a 1.2% increase in July. Overall headcount returned to slightly above pre-COVID levels in the middle of June[2], but the growth rate has slowed down significantly: June saw a 2.5% increase in net headcount compared to the prior month, and May recorded a 3.4% jump. Additionally, overall headcount is -78% lower that where it would be expected when comparing growth rates from previous years.
The rate of employer actions that affect employee wages (furloughs, terminations, reductions in hours and wages) continue to drop compared to last month. Year-over-year comparisons are returning to normal.
Table 1. Rates of employer actions affecting employee wages for 2019 and 2020 as compared to February of the same year[3].
An additional trend to note is that the frequency of workers who experienced a reduction in hours of at least 10% continues to be lower in July ‘20 than July ‘19. This is likely because the hourly workers who are generally most susceptible to having their hours reduced (this does not typically apply to salaried workers) were much more likely to simply be furloughed or terminated this year than ever before. Because of the way we measure employment status on a monthly cadence, furloughed or terminated employees cannot also be measured as “reduced hours of at least 10%” in the same month that they are unemployed, and also cannot be measured in any other category after they have been terminated.
Furloughs, reduction in hours, and reduction in hourly pay rates were near equal contributors to the overall reduction of total wages in July. Furloughs accounted for 29% of the total reduction in wages experienced by workers in July ‘20, reduction in hours contributed another 29%, and reduction in hourly wages accounted for 26%.
Hiring Slows
Overall small business headcount growth in May and June was driven mostly by recovered hiring. The same is true for July, although the pace of increased hiring has slowed noticeably since last month. Figure 1 shows monthly headcount statistics since January 2019. The hiring rate in July ‘20 (6.4%) indicates a deceleration compared to previous months (7.5% for June ‘20 and 7.3% for May ‘20 respectively). July ‘20 hiring is down -4% relative to the same month last year, and is -13% lower than the observed hiring rate last month.
While hiring is softening, net headcount change was also bolstered by lower than usual terminations for this time of year. The rate of termination in July ‘20 from the previous month was 5.1%, which is -7% lower than the same time last year. Terminations are likely lower than normal for several reasons. First off, in order to have PPP loans forgiven, employers are required to continue to use at least 60% of their loan on payroll. We’re also seeing data indicating that employers are already operating with reduced staff. It’s possible that some small businesses have simply reached the minimum number of employees that are required in order to actually operate their business.
Figure 1. Monthly net change in headcount and hiring, termination, and layoff statistics.
Rehiring Rate Approaches pre-COVID Baseline
The proportion of all hires that were re-hires spiked in May, likely due to PPP loan distribution. This rehiring rate has dropped since, and the proportion of all hires that are re-hires (plotted over the past year in Figure 2 below) is now approaching pre-COVID levels (~10% of all hires). This would indicate that if an employee was terminated in the midst of the strictest lockdowns in March and April and hasn’t been hired back already, they probably will not be rehired by that employer.
Figure 2. Re-hires as a percentage of hires, on a weekly basis.
Figure 3. Weekly change in headcount and hiring, termination, and layoff statistics.
Quitting Rates Back in Normal Range
The voluntary termination rate, an indicator that employees felt more comfortable taking new job opportunities and that there were some openings available, continued to tick marginally upwards in July ‘20 and is now in the same range as pre-COVID months. The month-over-month rate of increase has slowed, however, as July ‘20 is only 4% higher than June’s observed quit rate. For comparison, June saw 3.4% of employees leave their company voluntarily, which is a 34% increase over May.
Furloughs Fall For Third Straight Month
The percentage of employees temporarily furloughed fell for a third time in July, from 6.1% in June to 5.9%. This month’s decline in furloughs is not as steep as the past two months, signaling that the recovery towards normal employment may be slowing down. Furloughs remain unseasonably high, at +39% above the same period last year.
Furlough rates for salaried employees are holding steady at just above 2% (2.2% in July, 2.1% in May, and 2.4% in June). Hourly workers remain much more likely to be furloughed than salaried workers (8.7% for hourly workers in July, +306% higher than salaried workers), though the gap has narrowed since April and May. The left column of charts in Figure 4 shows the various changes in employment broken down by FLSA status, and the right column shows the year-over-year percent change from the same months last year. The accompanying tables for Figure 4 are provided in the appendix.
Figure 4: Percent of workers that experienced the following changes in employment as compared to February of the same year, cut by FLSA Status (salaried vs. hourly).
The way small businesses reduce wages varies between hourly and salaried workers. Salaried employees are still 3.5 times more likely than hourly workers to have their wages cut, while hourly workers are more than 4 times more likely than salaried workers to be furloughed and 3.5 times more likely to have had a reduction in hours of more than 10% in the month of July than salaried workers.
The rate of employees who experienced a reduction in hours was lower this July than the same time last year (-23% across both exempt and nonexempt employees compared to July ‘19), a phenomenon that can be explained by the persistent higher than normal rate of furloughs and terminations. As noted previously, because of the way we measure employment status on a monthly cadence, furloughed or terminated employees cannot also be measured as “reduced hours of at least 10%” in the same month that they are unemployed, and also cannot be measured in any other category after they have been terminated. Another possible explanation for this trend is that a reduction in total staff could mean less flexible work hours for employees. To keep a business operating but with fewer employees, an employer may not have the option to reduce hours.
Fewer Than Half of Furloughed Employees Have Returned to Work
For employees whose furlough began between March ‘20 and June ‘20, only 40% have subsequently returned to work. Depending on the month the employee was originally furloughed, however, return-to-work rates range from 37% (employees furloughed in March ‘20) to 47% (employees furloughed in April ‘20).
Historically, termination has been a more likely outcome of a furlough than we’ve seen in 2020. 22% of all employees furloughed between March ‘20 and June ‘20 have been terminated by the end of July ‘20, whereas the termination rate for the same time period last year was 31%. Furloughs in 2020 were a direct reaction to government lockdowns, and for the most part appear to be slightly more “temporary” than furloughs last year. On the flip side, employees furloughed this year have been much more likely to stay furloughed (28% as opposed to 21% in 2019), and while a similar proportion have returned to work compared to last year (40% versus 41%), those who returned to work have been much more likely to return to lower wages than they received before their furlough began (either because of reduced hours or reduced hourly rate). Of those that have returned to work this year, 34% have returned to lower wages, as opposed to 23% for the same time in 2019.
Table 2. Rate of Employer Actions Impacting Employees Furloughed Between March and June by July of the same year
Reviewing 2020 furlough data on a monthly cohort basis reveals that the timing of your furlough had a meaningful impact on the outcome of that furlough. Among employees furloughed since February ‘20, those furloughed in March have fared the worst. A full third of employees who were furloughed in March have now been terminated, and only 37% have returned to work. An even smaller number of those workers that have returned to work have returned to the same wages they were receiving prior to being furloughed (meaning they either had their hours reduced or their hourly pay rate cut upon return). While 37% of employees that were furloughed in March have returned to work by July, nearly 25% of those employees have returned to reduced wages. For comparison, 41% of employees that were furloughed in March of ‘19 had returned to work by July ‘19. Of those, only 21% returned at reduced wages.
Employees furloughed in April have experienced the most positive outcomes. As of July, 47.5% are back to work, and 34% of all furloughed employees are back at a wage comparable to their pre-pandemic baseline (28% of those employees who are back at work are working on reduced wages).
The rate of furloughs lasting through July ‘20 ranged from 19% (furloughed in March) to 34% (furloughed in June). This is in part explained by the duration of time each monthly cohort has had to reach July; as time goes on, the June cohort of remaining furloughed employees will go down as employees are either terminated or rehired. In July of last year, the rate of remaining on furlough ranged from 15% (furloughed in March) to 24% (furloughed in June).
Table 3. Rate of Employer Actions Impacting Cohorts of Furloughed Employees in Subsequent Months of 2020
Highest Salaried Earners Most Likely to Have Reduction in Wages
Salaried workers have been less impacted by headcount reductions than hourly workers in 2020, but there have still been changes to their employment in the face of a rocky economy. Reduction to earnings (average income reduction of at least 10%), while still a relatively uncommon occurrence, was 3.5x more prevalent for salaried workers than for hourly workers. Hourly workers were much more likely to be furloughed or terminated than to have their pay or hours adjusted.
The left column of charts in Figure 5 shows the various changes in employment broken out by income bracket for salaried workers, and the right column shows the year-over-year percent change from the same months last year. The accompanying tables for Figure 5 are provided in the appendix.
Salaried workers earning more than $100,000 a year have been the most likely to experience a reduction to their average hourly wage. 5.6% of these workers in this income bracket in July ‘20 were working at a wage that was at least 10% less than their earnings in February. While this is lower than the peak in May ‘20 (7.0%, which was +143% higher than May ‘19), July wage reduction for these high earners is still +91% higher than the same month last year.
While all salaried workers were more likely to experience a reduction in earnings since February compared to the same period last year, the lower a salaried worker’s income, the less likely they are to have experienced a reduction in earnings. Only 4.4% of salaried workers earning less than $50,000 per year had a reduction in their average hourly wage, although this is still +22% higher than the same time last year.
Figure 5. Rate of Employer Actions Impacting Salaried Employees since February ‘20
Highest Paid Hourly Workers Still Seeing Atypical Termination Patterns Compared to Last Year
Hourly workers (non-exempt FLSA status) making more than $30 per hour have been the least likely of all hourly workers to be terminated between February and July (21.5%, compared to 30.1% of workers earning less than $15 an hour), but their year-over-year percent change in termination rates compared to the same months of 2019 have been the highest of all hourly workers. Proportionally, these workers are seeing the most unexpected jumps in terminations due to COVID-related business disruption.
The first column of charts in Figure 6 shows the various changes in employment broken out by hourly wages for Non-Exempt employees, and the second column shows the year-over-year percent change from the same months last year. The accompanying tables for Figure 6 are provided in the appendix.
Workers making more than $30 per hour were terminated by July ‘20 at a rate +36% higher than July of last year. By July ‘20, workers making $15 an hour or less, on the other hand, have returned to normal termination rates compared to the same period last year (only +5% higher than the same period last year). In general, lower income hourly jobs tend to have more turnover than higher paying occupations, so while the termination rates for workers making $15 or less per hour in 2020 is the highest of all hourly wage brackets, it is also most quickly converging towards 0% difference compared to 2019 (as shown in the bottom right hand chart in Figure 6 below).
Figure 6. Rate of Employer Actions Impacting Hourly Employees Since February ‘20
Youngest Workers Most Likely to Be Unemployed
Headcount is below where we’d expect it to be in each age bracket, but is furthest behind for workers younger than 25. Headcount (the net combination of terminations and hiring) for workers younger than 20 is still -14.9% below where we’d expect it to be (using 2019 as a baseline), and -12.5% behind for workers 20-24 years old. Headcount of employees between 25 and 34 is -8.1% below expectations in July, and all age segments 35 and older were within -5% from the previous year headcount baseline.
Tables 4 and 5 show the ratio of monthly headcount volume in a given month to the headcount in February of the same year for 2020 and 2019, respectively. For instance, the March column of Table 4 is calculated by dividing the number of employees working in March 2020 by the number of employees in February 2020. If there were 1,000 employees 55 years or older in February ‘20, this table would indicate that there were 964 employees in that same age segment in March ‘20.
Table 6 then shows the comparison of 2020 monthly headcount changes versus 2019 as a baseline for what we believe would have happened if there were no pandemic and economic crisis. For instance, the relative percent difference between July ‘20 and July ‘19 is -14.9%.
Table 4. Headcount volume change in 2020 months compared to February
Table 5. Headcount volume change in 2019 months compared to February
Table 6. Year over Year comparison of 2020 (Table 3) to 2019 (Table 4)
Younger workers are more likely to be employed in low-salary or hourly roles where furloughs and terminations have been more prevalent and where turnover in general tends to be higher (see Figures 5 and 6). Given the prevalence of terminations and furloughs for the youngest workers, they were also therefore the least likely age segment to have their hourly rate reduced (19 or younger and 20-24). These workers also saw lower year-over-year change in the reduction of their hours than older workers, yet these differences converged slightly in July. Older workers were less likely to be terminated or furloughed, but more likely to have their pay reduced.
The first column of charts in Figure 7 shows the various changes in employment broken out by employee age, and the second column shows the year-over-year percent change from the same months last year. The accompanying tables for Figure 7 are provided in the appendix.
Figure 7. Rate of Employer Actions Impacting Employees since February ‘20, broken out by Employee Age
Industry Trends
Small business headcount grew across nearly all industries in July, though the pace of growth has slowed compared to the rapid gains made in May and June. A few industries saw a reduction in headcount in July compared to June (Sports & Recreation, Accounting), the first time that’s happened in any industry since April. Still, on the whole, companies across industries are growing. Some of the largest gains were made in the following industries:
- Transportation (+3.6%)
- Healthcare & Social Assistance (+2.5%)
- Retail (+1.9%)
Figure 8 shows net headcount changes for all industries with 1,000 Gusto customers over the past month, ranked by the cumulative change in headcount over the period.
Figure 8. Net change in headcount for March through June by industry.
Hardest-Hit Industries Are Still Operating on Low Headcount
Headcount growth in industries that have been hit hardest by the pandemic has plateaued over the last month, whereas other industries have continued to grow. Accommodations, Arts & Entertainment, Food & Bev, Salon & Spa, Sports and Recreation and Tourism have all maintained relatively flat staffing levels since June and are still operating with fewer employees than they were at the beginning of March.
Figure 9 shows the cumulative change in headcount by week since the first week of March 2020. As of the final week of July, Retail is +3.3% above its headcount in early March, as is the catch-all bucket for industries least impacted by COVID. Accomodations was on a steep line of recovery, but over the course of July decreased slightly and is now still -3.5% below where headcount was in early March. Salon & Spa has plateaued at -4% from its pre-COVID baseline; Sports & Recreation is at -7.5%, Food and Beverage is at -7.7%, Arts & Entertainment is at -8.6% and Tourism is at -28.7%.
This flattened trend seems to be due to a slowdown in hiring for these industries rather than an uptick in terminations. Following high hiring rates in May and June, July hiring rates fell across these specified industries while remaining mostly flat for “All Other”. The deep cuts to headcount that were made in March and April were unprecedented, and hiring would have needed to maintain reflectively high levels for several more months to make up the deficit. By July, termination rates across industries are in the same range of the same period last year.
Figure 9. Cumulative change in headcount since the week of March 2, 2020 across COVID-heavily affected industries and all others.
To more fully understand headcount changes, we’ve also used 2019 data as a baseline for what we expect would have happened if COVID of benchmarking to the first week of March of 2020 alone. Figure 10 below shows the cumulative change in headcount by week between the first week of March 2019 and the end of July 2019, and further clarifies that the trends we’re seeing in 2020 are highly abnormal. In a typical year not wracked by a global pandemic, overall headcount for companies that remain in business does not stay static at 0% month over month, but, on average, grows as time goes on. This comparison of these two graphs puts into stark relief just how far behind 2020 is for cumulative change in headcount thus far.
Figure 10. Prior Year cumulative change in headcount (between March 4, 2019 and July 28, 2019) across COVID-heavily affected industries and all others[5].
While overall headcount is now slightly positive (+2.0%) compared to the end of February ‘20, it’s still -78% below where we expect it would be based on 2019 data. For highlighted industries (laid out in Table 7), it’s even worse. Table 6 below shows the comparison of each industry’s February ‘20 to July ‘20 growth rate versus the February ‘19 to July
Looking at these monthly cumulative headcount change, Food & Beverage in 2020 is slightly less than -7%, but compared to 2019’s cumulative headcount change between February and July of +13.4, we can conclude that Food & Beverage headcount is -150% where it would have been if not for the COVID-19 crisis. Salon & Spa (-179%), Sports & Recreation (-152%), Arts & Entertainment (-229%) and Tourism (-304%) show even worse year over year comparisons.
Table 7. Year over Year Comparison of cumulative change in headcount from end of February through July in 2020 and 2019.
Geographic Trends
Similar to June, nearly all US states with at least 1,000 employees on the Gusto platform also showed net increases in headcount for the month of July ‘20. The only exceptions were North Dakota, Wyoming, which showed very slight declines, and Iowa, which saw a moderate decline of -2.3%.
Figure 11. Net percentage change in headcount for March–May by state.
About Gusto
Gusto is a modern, online people platform that helps small businesses take care of their teams. In addition to full-service payroll, Gusto offers health insurance, 401(k)s, compliance and expert HR, and more. The company serves over 100,000 businesses nationwide and has offices in Denver, New York City, and San Francisco.
Appendix
Table A1: Hires, terminations, and layoffs per 100 employees by industry, for the month of July ‘20, ranked in decreasing order by net change in headcount.
| Industry | Hires | Terminations | Layoffs | % Change in hiring vs. march | Percentage change in terminations | Percentage change in layoffs | Net change in employees |
| Facilities | 14.47 | 8.14 | 0.47 | 0.3 | -0.12 | -0.74 | 6.33 |
| Food & Beverage | 13.78 | 9.17 | 2.07 | 1.33 | -0.07 | -0.47 | 4.61 |
| Construction | 8.58 | 4.23 | 0.49 | 0.52 | -0.21 | -0.65 | 4.35 |
| Transportation | 10.35 | 6.27 | 0.78 | 0.44 | -0.2 | -0.68 | 4.08 |
| Retail | 9.17 | 5.15 | 0.7 | 0.46 | -0.16 | -0.63 | 4.02 |
| Manufacturing | 7.68 | 4.03 | 0.68 | 0.54 | -0.19 | -0.65 | 3.64 |
| Wholesale | 6.98 | 3.38 | 0.91 | 0.65 | -0.17 | -0.48 | 3.6 |
| Healthcare & Social Assistance | 7.59 | 4.08 | 0.3 | 0.88 | -0.15 | -0.79 | 3.51 |
| Insurance | 5.63 | 2.27 | 0.2 | 0.54 | -0.35 | -0.75 | 3.37 |
| Salon & Spa | 10.48 | 7.24 | 1.49 | 3.61 | 0.58 | -0.11 | 3.24 |
| Consulting | 6.06 | 3.01 | 0.64 | 0.42 | -0.24 | -0.54 | 3.04 |
| Other Personal Services | 7.65 | 4.69 | 0.71 | 1.09 | -0.31 | -0.78 | 2.96 |
| Education | 6.38 | 3.6 | 0.44 | 0.97 | -0.3 | -0.74 | 2.78 |
| Other Professional Services | 7.46 | 4.72 | 0.59 | 0.63 | -0.2 | -0.73 | 2.74 |
| Communications | 4.8 | 2.1 | 0.44 | 0.79 | -0.58 | -0.86 | 2.7 |
| Real Estate | 5.43 | 2.83 | 0.38 | 0.68 | -0.36 | -0.68 | 2.6 |
| Technology | 4.64 | 2.17 | 0.35 | 0.39 | -0.34 | -0.71 | 2.48 |
| Finance | 4.44 | 1.97 | 0.37 | 0.3 | -0.39 | -0.67 | 2.47 |
| Legal | 4.68 | 2.26 | 0.3 | 0.83 | -0.27 | -0.6 | 2.41 |
| Sports & Recreation | 8.23 | 6.04 | 1.3 | 1.54 | -0.26 | -0.56 | 2.19 |
| Non-Profits & Associations | 3.75 | 2.02 | 0.23 | 0.69 | -0.49 | -0.79 | 1.74 |
| Accounting | 4.63 | 3.01 | 0.47 | 0.91 | -0.41 | -0.78 | 1.62 |
| Arts & Entertainment | 4.76 | 4.02 | 1.1 | 0.87 | -0.46 | -0.62 | 0.74 |
Table A2: Hires, terminations, and layoffs per 100 employees across the most populous 50 Metropolitan Statistical Areas for the month of July ‘20, ranked in decreasing order by net change in headcount.
| msa_title | Hires | Terminations | Layoffs | Percentage change in hires | Percentage change in terminations | Percentage change in layoffs | Net change in employees |
| Louisville/Jefferson County, KY-IN | 8.72 | 2.87 | 0.43 | 0.62 | -0.2 | -0.86 | 5.85 |
| Columbus, OH | 10.21 | 4.47 | 0.22 | 0.97 | -0.27 | -0.81 | 5.74 |
| Houston-The Woodlands-Sugar Land, TX | 9.94 | 4.55 | 0.5 | 1.39 | -0.18 | -0.71 | 5.39 |
| Detroit-Warren-Dearborn, MI | 9.18 | 3.85 | 0.65 | 1.18 | -0.34 | -0.81 | 5.33 |
| Cleveland-Elyria, OH | 9.16 | 3.88 | 0.07 | 0.22 | 0.09 | -0.92 | 5.28 |
| Kansas City, MO-KS | 8.06 | 3.16 | 0.26 | 0.5 | -0.38 | -0.73 | 4.9 |
| Hartford-West Hartford-East Hartford, CT | 7.94 | 3.12 | 0.51 | 0.96 | -0.26 | 0.17 | 4.81 |
| Salt Lake City, UT | 8.61 | 4.14 | 0.34 | 0.6 | -0.24 | -0.82 | 4.47 |
| Milwaukee-Waukesha-West Allis, WI | 8.89 | 4.51 | 0.22 | 1.42 | -0.14 | -0.82 | 4.38 |
| Tampa-St. Petersburg-Clearwater, FL | 9.54 | 5.17 | 0.47 | 0.66 | -0.1 | -0.65 | 4.37 |
| Providence-Warwick, RI-MA | 8.29 | 3.95 | 0.2 | 1.2 | -0.19 | -0.9 | 4.34 |
| Jacksonville, FL | 8.75 | 4.51 | 0.16 | 0.91 | -0.26 | -0.92 | 4.24 |
| Austin-Round Rock, TX | 7.95 | 3.98 | 0.53 | 0.92 | -0.33 | -0.78 | 3.97 |
| Dallas-Fort Worth-Arlington, TX | 9.34 | 5.4 | 0.45 | 0.78 | 0.03 | -0.69 | 3.94 |
| Atlanta-Sandy Springs-Roswell, GA | 8.26 | 4.41 | 0.57 | 1.18 | 0.06 | -0.47 | 3.85 |
| San Antonio-New Braunfels, TX | 8.79 | 4.94 | 0.46 | 0.75 | -0.07 | -0.64 | 3.84 |
| Charlotte-Concord-Gastonia, NC-SC | 7.52 | 3.68 | 0.19 | 0.92 | -0.43 | -0.94 | 3.84 |
| Virginia Beach-Norfolk-Newport News, VA-NC | 6.12 | 2.36 | 0.09 | 0.16 | -0.38 | -0.87 | 3.77 |
| Miami-Fort Lauderdale-West Palm Beach, FL | 8.51 | 4.75 | 0.84 | 0.98 | -0.03 | -0.31 | 3.75 |
| Baltimore-Columbia-Towson, MD | 6.72 | 3.02 | 0.26 | 1.02 | -0.35 | -0.76 | 3.7 |
| Denver-Aurora-Lakewood, CO | 7.07 | 3.39 | 0.46 | 0.62 | -0.34 | -0.75 | 3.68 |
| Orlando-Kissimmee-Sanford, FL | 7.97 | 4.5 | 0.56 | 1.22 | -0.27 | -0.47 | 3.47 |
| Phoenix-Mesa-Scottsdale, AZ | 8.66 | 5.21 | 0.51 | 0.86 | -0.02 | -0.6 | 3.45 |
| Indianapolis-Carmel-Anderson, IN | 6.39 | 3 | 0.28 | 0.49 | -0.61 | -0.93 | 3.39 |
| Nashville-Davidson–Murfreesboro–Franklin, TN | 9.22 | 5.93 | 1.09 | 0.31 | 0 | -0.47 | 3.29 |
| Cincinnati, OH-KY-IN | 7.24 | 3.97 | 0.31 | 0.72 | 0.08 | -0.72 | 3.27 |
| Birmingham-Hoover, AL | 7.22 | 3.97 | 0.58 | 1.6 | 0.1 | 2.91 | 3.25 |
| New York-Newark-Jersey City, NY-NJ-PA | 7.24 | 4.01 | 0.86 | 1.1 | -0.38 | -0.7 | 3.22 |
| San Diego-Carlsbad, CA | 6.71 | 3.52 | 0.48 | 0.9 | -0.2 | -0.72 | 3.19 |
| Las Vegas-Henderson-Paradise, NV | 7.71 | 4.55 | 1.35 | 0.47 | -0.37 | -0.61 | 3.16 |
| New Orleans-Metairie, LA | 6.39 | 3.42 | 0.59 | 0.43 | -0.53 | -0.51 | 2.97 |
| Sacramento–Roseville–Arden-Arcade, CA | 6.73 | 3.82 | 0.45 | 0.88 | -0.22 | -0.62 | 2.9 |
| Portland-Vancouver-Hillsboro, OR-WA | 6.35 | 3.51 | 0.67 | 0.64 | -0.26 | -0.6 | 2.84 |
| Minneapolis-St. Paul-Bloomington, MN-WI | 6.46 | 3.67 | 0.68 | 0.62 | -0.38 | -0.43 | 2.79 |
| Los Angeles-Long Beach-Anaheim, CA | 6.26 | 3.58 | 0.73 | 0.74 | -0.19 | -0.61 | 2.68 |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 6.26 | 3.65 | 0.98 | 0.64 | -0.12 | -0.25 | 2.61 |
| St. Louis, MO-IL | 8.02 | 5.42 | 1.54 | 0.15 | -0.3 | -0.28 | 2.6 |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 6.77 | 4.19 | 0.42 | 0.88 | -0.19 | -0.81 | 2.58 |
| Boston-Cambridge-Newton, MA-NH | 6.03 | 3.56 | 0.7 | 0.5 | 0 | -0.41 | 2.47 |
| Chicago-Naperville-Elgin, IL-IN-WI | 5.67 | 3.27 | 0.73 | 0.41 | -0.38 | -0.63 | 2.4 |
| Pittsburgh, PA | 5.93 | 3.62 | 0.31 | 0.26 | -0.53 | -0.93 | 2.32 |
| San Francisco-Oakland-Hayward, CA | 5.33 | 3.34 | 0.89 | 0.5 | -0.31 | -0.52 | 1.98 |
| Memphis, TN-MS-AR | 8.85 | 6.89 | 0.15 | 0.31 | -0.02 | -0.93 | 1.96 |
| Richmond, VA | 6.61 | 4.69 | 0.36 | 0.75 | -0.18 | -0.73 | 1.93 |
| Seattle-Tacoma-Bellevue, WA | 6.01 | 4.17 | 0.8 | 0.66 | -0.38 | -0.7 | 1.84 |
| Raleigh, NC | 7.1 | 5.46 | 0.55 | 1.05 | 0.09 | -0.68 | 1.64 |
| San Jose-Sunnyvale-Santa Clara, CA | 5.01 | 3.6 | 0.82 | 0.85 | -0.31 | -0.66 | 1.41 |
| Riverside-San Bernardino-Ontario, CA | 5.96 | 4.61 | 0.98 | 0.53 | -0.09 | -0.6 | 1.35 |
| Buffalo-Cheektowaga-Niagara Falls, NY | 3.88 | 2.88 | 0.43 | 0.32 | -0.71 | -0.94 | 1.01 |
| Oklahoma City, OK | 7.94 | 8.51 | 0.7 | 0.71 | 0.48 | -0.5 | -0.57 |
Table A3: Percent of workers that experienced the following changes in employment as compared to February of the same year, cut by FLSA Status (salaried vs. hourly).
Table A4: Percent of workers that experienced the following changes in employment as compared to February of the same year, cut by Salary bracket.
Table A5: Percent of workers that experienced the following changes in employment as compared to February of the same year, cut by Hourly Wage bracket for nonexempt employees (hourly).
Table A6: Percent of workers that experienced the following changes in employment as compared to February of the same year, cut by Age bracket.
Methodology
The analyses described in this report were based on two separate datasets obtained from Gusto’s small business payroll, benefits, and HR applications. The first dataset used payroll data to capture employee hours, wages, and furloughing (Work Reduction & Furlough). The second dataset used data from employers hiring and terminating employees (Hiring & Termination). When we define outcomes, we’re assigning one outcome per employee per month, which can block certain outcomes (Outcome Rank Ordering).
Work Reduction & Furlough
Employees in our dataset were labeled as furloughed if they were employed for the entirety of February, March, April and May (had no termination effective during that time), and either:
- had earned money in the final paycheck of the preceding month but had not earned money in the final paycheck of the following month (e.g. final paycheck of February was non-zero, and the final paycheck of March was $0), or
- met the definition for part 1 and also continued to earn $0 on payrolls in the subsequent month, thereby on a continued furlough.
Our furlough data in this April report only looks at furloughs that were active in April and were instigated in March or April, not furloughs that began earlier and have continued through this period.
Non-terminated, non-furloughed employees were labeled as work reduced for the June report if the total hours of their June paychecks were less than 90% of the hours on their February (used as a benchmark for pre-COVID) paychecks. This calculation was done for both 2019 and 2020, where 2019 was used to represent the counterfactual case (what would have happened).
Presumed COVID-related differences were calculated by subtracting the rates of termination, furlough, and work reduction in 2019 from the same numbers in 2020. This rate difference was used to calculate the number of additional employees affected by each cost-saving strategy. The average dollars saved per affected employee was calculated as the difference between total February and March payrolls, divided by the number of employees, for each type of affected employee, and this dollar amount was multiplied by the COVID-related count of impacted employees to calculate the total dollars saved by each strategy.
Hiring & Termination
A given employee on Gusto can have multiple “employments,” since an employer can potentially hire, terminate, and rehire the same employee multiple times. In our Hiring & Termination dataset, a hire corresponded to an employer creating a new employment with a hiring date, and a termination corresponded to the entry of a termination date for a given employment. Layoffs corresponded to terminations where the employer listed the reason for the termination as a “layoff” (one of the choices from a standardized list in Gusto’s termination flow). In order to capture a more time-sensitive view of employer activities, we recorded hires, terminations, and layoffs on the date that they were entered into the system, rather than their effective date. Terminations are also coded as voluntary or involuntary.
Monthly reporting of termination data may differ slightly from prior reporting, as employers often log a termination several days or weeks after the employee was actually released. For example, more than a third of terminations that were created (logged in the system) in May were actually effective before May. Data in this report may also differ slightly from prior reports due to methodology improvements. Further employment changes are incorporated into this month’s analysis, such as reduction to hourly wages. We also separately categorized employees that worked for companies who “paused” their payroll service with Gusto. Employees that fall into this new bucket may have previously not been labeled as impacted, or may have been attributed as “hours reduced” for a given month.
When we assign an outcome to an employee for a given month, we follow this rank ordering:
- Employee Terminated
- Company Suspended/Paused
- Employee Furloughed
- Hours Reduced
- Hourly Rate Reduced
What this means, in practice, is that if an employee’s hourly rate was reduced early in the month and then they were furloughed, we’ll count them toward the “Furloughed” outcome but not the “Hourly Rate Reduced” outcome. The further down the rank order the outcome is, the more likely we are to undercount it in favor of something else.
Employee hires, layoffs, and terminations were aggregated weekly and monthly for a given company and work location. Hiring, termination, and layoff rates reported in this document are based on the number of times each event occurred in a given week or month, divided by the number of employees who were employed at the beginning of the period.
Locations reported in this document are based on the most recent work location associated with the employee. Industries reported in this document are based on self-report from customers within Gusto’s product.
[1] Values provided in this report were only included for categories (e.g., state, industry, metropolitan area) where we had at least 1,000 employees.
[2] An initial attempt to adjust for industry skew using NAICS codes showed adjusted headcount to have just crept above zero this month, but by less than half a percent.
[3] Historical data may differ from earlier Gusto reporting as employers may log employee terminations after the actual termination date (e.g. more than a third of terminations that were logged in April by employers actually occurred in a prior month).
[4] Employees in the Other bucket could have worked for companies that went out of business, left Gusto’s platform, etc.
[5] Accommodations is excluded as a standalone category in the 2019 plot due to an insufficiently large sample size for that time period.
