U.S. states ranked by working from home
Compare the share of workers who usually work from home in all 50 states. Uses 2020–2024 ACS commuting data with a worker denominator and clear limitations.
Which states have the largest work-from-home shares?
Colorado has the highest point estimate, 21.0%, for 2020–2024. Next are Washington (19.8%) and Maryland (19.3%).
Allowing for margins of error, Colorado’s first place is statistically clear.
Colorado leads the point estimates, followed by Washington. These figures describe a five-year period that includes major pandemic-era commuting changes, rather than remote-work prevalence in September 2026.
The denominator is workers age 16 and older. Working from home is the source-defined usual commuting category, not a count of everyone who occasionally answers email at home.
- Reference period
- 2020–2024
- Original source
-
U.S. Census Bureau · ACS 2024 five-year detailed tables, B08301
Source published 2026-01-29 - Ordering
- 50 states, highest exact point estimate first; no composite score
- Data and methods reviewed
- October 7, 2026 · data snapshot apd-2026-10-07-r5
Complete 50-state ranking table
Filter the list or change its display order. Positions always refer to the full 50-state ordering. The ± column contains the derived 90% margin of error. The supported range lists the positions that pairwise 90% significance tests cannot rule out; a single number means the position is statistically clear.
Showing all 50 states.
| Position | State | Percentage | 90% margin of error | Supported range |
|---|---|---|---|---|
| 1 | Colorado | 21.0% | ±0.2 pp | 1st |
| 2 | Washington | 19.8% | ±0.2 pp | 2nd |
| 3 | Maryland | 19.3% | ±0.2 pp | 3rd |
| 4 | Oregon | 19.0% | ±0.2 pp | 4th |
| 5 | Massachusetts | 18.6% | ±0.2 pp | 5th |
| 6 | Arizona | 18.4% | ±0.2 pp | 6th–7th |
| 7 | Virginia | 18.3% | ±0.2 pp | 6th–7th |
| 8 | Minnesota | 17.6% | ±0.1 pp | 8th–9th |
| 9 | Utah | 17.6% | ±0.3 pp | 8th–9th |
| 10 | California | 17.2% | ±0.1 pp | 10th–12th |
| 11 | Vermont | 16.9% | ±0.4 pp | 10th–13th |
| 12 | New Hampshire | 16.8% | ±0.4 pp | 10th–13th |
| 13 | New Jersey | 16.8% | ±0.2 pp | 11th–13th |
| 14 | North Carolina | 16.3% | ±0.2 pp | 14th |
| 15 | Connecticut | 16.0% | ±0.2 pp | 15th–18th |
| 16 | Georgia | 15.9% | ±0.1 pp | 15th–18th |
| 17 | Maine | 15.9% | ±0.4 pp | 15th–20th |
| 18 | Illinois | 15.9% | ±0.1 pp | 15th–18th |
| 19 | Florida | 15.7% | ±0.1 pp | 18th–20th |
| 20 | Pennsylvania | 15.6% | ±0.1 pp | 18th–20th |
| 21 | Delaware | 15.0% | ±0.5 pp | 21st–22nd |
| 22 | New York | 14.9% | ±0.1 pp | 21st–22nd |
| 23 | Texas | 14.4% | ±0.1 pp | 23rd |
| 24 | Michigan | 13.5% | ±0.1 pp | 24th–26th |
| 25 | Idaho | 13.3% | ±0.3 pp | 24th–29th |
| 26 | Wisconsin | 13.1% | ±0.1 pp | 25th–30th |
| 27 | Missouri | 13.0% | ±0.2 pp | 25th–30th |
| 28 | Rhode Island | 13.0% | ±0.5 pp | 24th–31st |
| 29 | Ohio | 13.0% | ±0.1 pp | 25th–30th |
| 30 | Tennessee | 12.9% | ±0.2 pp | 26th–31st |
| 31 | Montana | 12.5% | ±0.4 pp | 29th–32nd |
| 32 | New Mexico | 12.2% | ±0.3 pp | 31st–34th |
| 33 | Nevada | 11.9% | ±0.3 pp | 32nd–34th |
| 34 | Kansas | 11.8% | ±0.2 pp | 32nd–35th |
| 35 | South Carolina | 11.5% | ±0.2 pp | 34th–36th |
| 36 | Iowa | 11.3% | ±0.2 pp | 35th–37th |
| 37 | Nebraska | 11.0% | ±0.3 pp | 36th–37th |
| 38 | Indiana | 10.7% | ±0.2 pp | 38th |
| 39 | Kentucky | 10.2% | ±0.2 pp | 39th |
| 40 | Alaska | 9.7% | ±0.4 pp | 40th–44th |
| 41 | South Dakota | 9.7% | ±0.3 pp | 40th–44th |
| 42 | Hawaii | 9.6% | ±0.3 pp | 40th–44th |
| 43 | Wyoming | 9.4% | ±0.5 pp | 40th–47th |
| 44 | Oklahoma | 9.4% | ±0.2 pp | 40th–45th |
| 45 | West Virginia | 9.1% | ±0.3 pp | 43rd–47th |
| 46 | Alabama | 8.8% | ±0.2 pp | 44th–47th |
| 47 | Arkansas | 8.8% | ±0.2 pp | 44th–47th |
| 48 | North Dakota | 8.1% | ±0.3 pp | 48th–49th |
| 49 | Louisiana | 8.1% | ±0.2 pp | 48th–49th |
| 50 | Mississippi | 5.9% | ±0.2 pp | 50th |
The downloads preserve exact values, IDs, source counts, uncertainty status, source release and original ordering. The CSV is the full 50-state table, including when the on-page filter hides rows.
National benchmark and District of Columbia
The United States value is 15.1%. The District of Columbia value is 33.0%. Both use the same source product as the state table. They do not receive a numbered state position. The national estimate is a source value or an approved calculation from national source counts; it is not an average of the state medians or percentages.
What the lower end of this table means
Mississippi has the lowest point estimate, 5.9%. The last three entries are North Dakota (8.1%), Louisiana (8.1%), Mississippi (5.9%) . This ordering is specific to working from home and 2020–2024; it is not a verdict on the state’s overall living conditions.
Usual commute is not a hybrid-work schedule
The ACS commuting question describes how a worker usually gets to work. A hybrid worker may appear in a travel category depending on the reported usual pattern. This dataset does not supply days per week at home, remote-work eligibility or employer policies.
The survey period spans changing working patterns
Pooling 2020–2024 provides a common product across state and county profiles. It also averages across a period of substantial changes. This ranking cannot be labeled a current-year remote-work rate or used as an annual trend. Its frozen definition makes cross-state context reproducible without pretending to measure today’s arrangements.
Definition, calculation and uncertainty
Workers age 16 and older who usually worked from home as a share of all workers age 16 and older. The survey describes the usual journey to work. It is not a complete measure of hybrid work or all travel.
100 × (B08301_021E) / (B08301_001E)
The derived margins of error use the Census Bureau’s approximation formulas for proportions, applied to the published margins of error of the source counts. Two states are treated as different only when their difference exceeds the square root of the sum of their squared margins of error, a 90% test. The supported range runs from one plus the number of states significantly above to 50 minus the number significantly below. Tests are pairwise and not adjusted for multiple comparisons. Sorting uses full stored precision. Exact equal values share a position; the next position skips the tied places. A missing, suppressed or unavailable estimate would be excluded and identified rather than converted to zero.
Sources and reproducibility
Original source table and documentation · Census guidance on comparing ACS estimates · Population Estimates Program methods .
Editorial ordering release apd-editorial-2026-10-07-r6 uses the unchanged data snapshot apd-2026-10-07-r5. Cite the Census source product, table, 2020–2024, the indicator definition and this page’s snapshot. The retained JSON query makes the inclusion rule, calculation and sort direction explicit. An upstream revision will require a new approved release rather than rewriting this result.