How Many Jobs Can Actually be Done at Home?

The answer is 37%!

The pandemic brought to the forefront the reality of working from home (WFH). We’ve been part of many discussions regarding different aspects of WFH, among others the impact on the environment, productivity, and employee experience. One fundamental question that kept lingering in my mind, is how many of us can actually work from home?

It turns out that this question was answered in the scientific article “How many jobs can be done at home?” published a few months after the pandemic started, in September 2020.

Here is how they performed their study.

Which jobs can be done from home?

The authors used a public library or around 1000 occupations and surveyed on average 25 people for each one of them. Each respondent answered two questionnaires to determine if their job  cannot be done at home. 

The first one determined “the physical and social factors that influence the nature of work” and if any of the following conditions was true, the job cannot  be done at home:

  1. Average respondent says they use email less than once per month
  2. Average respondent says they deal with violent people at least once a week
  3. Majority of respondents say they work outdoors every day
  4. Average respondent says they are exposed to diseases or infection at least once a week
  5. Average respondent says they are exposed to minor burns, cuts, bites, or stings at least once a week
  6. Average respondent says they spent majority of time walking or running
  7. Average respondent says they spent majority of time wearing common or specialized protective or safety equipment

The second survey captures “the general types of job behaviors occurring on multiple jobs” and if any of the following conditions was true, the job cannot  be done at home:

  1. Performing General Physical Activities is very important
  2. Handling and Moving Objects is very important
  3. Controlling Machines and Processes [not computers nor vehicles] is very important
  4. Operating Vehicles, Mechanized Devices, or Equipment is very important
  5. Performing for or Working Directly with the Public is very important
  6. Repairing and Maintaining Mechanical Equipment is very important
  7. Repairing and Maintaining Electronic Equipment is very important
  8. Inspecting Equipment, Structures, or Materials is very important


The authors describe in their paper the exact methodology to measure and determine how many jobs can be performed at home, but their conclusion is “ We find that 37% of jobs in the United States can be performed entirely at home, with significant variation across cities and industries.” 

The following table gives the percentage of jobs in each industry that can be done at home:

Table 3. Share of jobs that can be done at home, by industry.

Table 3UnweightedWeighted by wage
Educational Services0.830.71
Professional, Scientific, and Technical Services0.800.86
Management of Companies and Enterprises0.790.86
Finance and Insurance0.760.85
Wholesale Trade0.520.67
Real Estate and Rental and Leasing0.420.54
Federal, State, and Local Government0.410.47
Other Services (except Public Administration)0.310.43
Administrative and Support and Waste Management and Remediation Services0.310.43
Arts, Entertainment, and Recreation0.300.36
Mining, Quarrying, and Oil and Gas Extraction0.250.37
Health Care and Social Assistance0.250.24
Transportation and Warehousing0.190.25
Retail Trade0.140.22
Agriculture, Forestry, Fishing and Hunting0.080.13
Accommodation and Food Services0.040.07

As is expected, desk jobs can be done at home at a far higher percentage than manual jobs.

Another interesting conclusion is that the 37% of jobs account for 46% of all wages, which means that, on average, WFH jobs not only were maintained during the pandemic, but they were paying better than the rest.

The demographics of the study show that metropolitan areas tend to have more jobs that can be done from home: “Across all metropolitan areas, the share of jobs that can be performed at home is strongly positively correlated with median household income (0.53) and its share of residents who attained a college degree (0.71) and negatively correlated with its home ownership rate (−0.31) and its share of residents who are white (−0.12)”

Table 2. Share of jobs that can be done at home, by metropolitan area.

Table 2Share of jobsMetropolitan characteristics
UnweightedWeighted by wageBA shareMedian incomeWhite shareOwner share
Top ten
San Jose-Sunnyvale-Santa Clara, CA0.510.660.501150.460.57
Washington-Arlington-Alexandria, DC-VA-MD-WV0.500.640.511010.540.63
Durham-Chapel Hill, NC0.460.570.47600.620.60
Austin-Round Rock, TX0.460.580.44730.770.58
San Francisco-Oakland-Hayward, CA0.450.580.491000.500.54
Boston-Cambridge-Nashua, MA-NH0.440.550.47860.760.62
Bridgeport-Stamford-Norwalk, CT0.440.580.47930.730.67
Hartford-West Hartford-East Hartford, CT0.440.530.39760.760.67
Salt Lake City, UT0.430.530.34710.800.67
Des Moines-West Des Moines, IA0.430.530.37690.870.69
Bottom ten
Baton Rouge, LA0.300.360.28570.590.68
Las Vegas-Henderson-Paradise, NV0.300.370.24570.610.53
Riverside-San Bernardino-Ontario, CA0.300.350.21620.610.63
Scranton–Wilkes-Barre–Hazleton, PA0.300.360.25520.900.68
McAllen-Edinburg-Mission, TX0.300.310.18380.880.68
Grand Rapids-Wyoming, MI0.290.370.32610.840.73
Lancaster, PA0.290.360.27640.890.68
Bakersfield, CA0.290.360.16520.750.58
Stockton-Lodi, CA0.290.330.18610.560.56
Cape Coral-Fort Myers, FL0.280.340.28550.850.71

Outside of the United States

Over the border the results are similar for countries with similar economies like the US: advanced economies with a well developed services industry. In these types of countries, the percentage of jobs that could be done at home was in the same ballpark of 40%. And there is a direct correlation between GDP and this figure: in Mexico and Turkey fewer than 25% of jobs can be done at home, while in Sweden and the United Kingdom that goes above 40%.

The conclusion of these figures is that countries with weaker economies had even fewer jobs that could be done at home during the pandemic.

Although we are often intimidated by scientific papers, this one is relatively easy to read even for non experts in the field, and I hope that I did justice to its main findings. If you want to dig in the data, they have even made everything available on GitHub.