A guide to NTIA / U.S. Census Bureau data
How to read U.S. state social-network-use estimates
Find a state estimate. Understand its scope and how much sampling uncertainty is reported.
Read the percentage together with its standard error. NTIA publishes 51 geographic estimates—50 states plus D.C.—for the same survey question. They support a scoped view of participation; the percentages alone do not establish a definitive ranking of the underlying populations.
United States · Internet users aged 15+ within each state or D.C. · November 2023 · At least occasional activity in the prior six months
Source: NTIA Internet Use Survey, administered by the U.S. Census Bureau. These are people reporting social-network activity, rather than accounts, all residents or users of a specific platform. Open the alphabetical table ↓
The uncertainty belongs beside the estimate
For illustration, California has a reported share of 73.5%, with standard error 0.89 pp. Wyoming's share is 76.4%, with standard error 1.97 pp. The chart below shows the errors themselves; longer bars indicate larger reported sampling uncertainty.
Source: NTIA Analyze Table, Nov 2023 / socialNetworkUser / adultInternetUser; CAPropSE and WYPropSE, multiplied by 100. Social Usage presentation. These bars show source standard errors, not participation percentages or confidence intervals.
Read the two example estimates and errors
| Geography | Share (%) | Source SE (pp) |
|---|---|---|
| California | 73.4996 | 0.8908 |
| Wyoming | 76.3998 | 1.9701 |
Source: NTIA aggregate CSV, retained row 769 including the header; CAProp / CAPropSE and WYProp / WYPropSE. Two named illustrative states with different source standard errors. They are not selected as extrema of participation or precision and do not represent a complete state ranking.
A standard error describes sampling variability in the estimated percentage. Its unit here is percentage points. It is not a participation rate or a claim that every survey error lies within that amount. The 0.89 pp value is also not a 95% margin of error.
Definitions: NTIA table code and dictionary. The retained official Explorer code uses a separate 1.96 × SE conversion for displayed 95% margins. This guide reports the original SEs and calculates no confidence intervals.
Find your state in the full table
The table is alphabetical, includes D.C. and keeps source precision. Every percentage has its own within-geography internet-user denominator. The explorer and permitted CSV provide the same fixed source release.
Open all 51 geographic estimates and source errors
| Geography | Share (%) | Source SE (pp) | Source flags |
|---|---|---|---|
| Alabama | 77.8767 | 1.6798 | None recorded |
| Alaska | 76.8998 | 2.1618 | None recorded |
| Arizona | 69.5264 | 2.3253 | None recorded |
| Arkansas | 79.7856 | 1.5336 | None recorded |
| California | 73.4996 | 0.8908 | None recorded |
| Colorado | 79.1929 | 2.2947 | None recorded |
| Connecticut | 77.5818 | 1.9816 | None recorded |
| Delaware | 74.4333 | 2.2228 | None recorded |
| District of Columbia | 81.3688 | 1.4360 | None recorded |
| Florida | 78.7755 | 1.0640 | None recorded |
| Georgia | 77.9015 | 1.8434 | None recorded |
| Hawaii | 65.3704 | 2.3268 | None recorded |
| Idaho | 77.5391 | 2.0313 | None recorded |
| Illinois | 76.4557 | 1.4885 | None recorded |
| Indiana | 78.5287 | 1.6230 | None recorded |
| Iowa | 81.1160 | 2.1679 | None recorded |
| Kansas | 79.6729 | 2.1360 | None recorded |
| Kentucky | 76.4678 | 1.9991 | None recorded |
| Louisiana | 73.6249 | 1.7175 | None recorded |
| Maine | 77.1907 | 1.9800 | None recorded |
| Maryland | 78.0648 | 2.2096 | None recorded |
| Massachusetts | 74.3224 | 1.5296 | None recorded |
| Michigan | 76.5942 | 1.3313 | None recorded |
| Minnesota | 79.7777 | 1.6285 | None recorded |
| Mississippi | 78.3358 | 2.0815 | None recorded |
| Missouri | 79.4689 | 1.6652 | None recorded |
| Montana | 73.7230 | 2.2376 | None recorded |
| Nebraska | 80.9216 | 2.0103 | None recorded |
| Nevada | 72.9598 | 2.1860 | None recorded |
| New Hampshire | 77.5931 | 1.6137 | None recorded |
| New Jersey | 74.2156 | 1.6343 | None recorded |
| New Mexico | 72.7012 | 1.6696 | None recorded |
| New York | 75.7233 | 1.2244 | None recorded |
| North Carolina | 77.9419 | 1.5851 | None recorded |
| North Dakota | 75.1476 | 2.2934 | None recorded |
| Ohio | 76.8913 | 1.3545 | None recorded |
| Oklahoma | 80.2466 | 1.7651 | None recorded |
| Oregon | 75.7467 | 1.5102 | None recorded |
| Pennsylvania | 74.8731 | 1.4130 | None recorded |
| Rhode Island | 74.1444 | 2.4605 | None recorded |
| South Carolina | 74.7718 | 1.7047 | None recorded |
| South Dakota | 80.9555 | 2.0477 | None recorded |
| Tennessee | 74.2756 | 1.4906 | None recorded |
| Texas | 79.1173 | 0.8975 | None recorded |
| Utah | 78.6973 | 1.8914 | None recorded |
| Vermont | 74.6691 | 1.9200 | None recorded |
| Virginia | 73.3118 | 2.0353 | None recorded |
| Washington | 70.9898 | 1.8531 | None recorded |
| West Virginia | 76.6783 | 1.5510 | None recorded |
| Wisconsin | 77.5057 | 1.6874 | None recorded |
| Wyoming | 76.3998 | 1.9701 | None recorded |
Source: NTIA Analyze Table, Nov 2023 / socialNetworkUser / adultInternetUser, retained row 769 including the header. Each state's postal abbreviation identifies its Prop and PropSE columns; D.C. uses DCProp / DCPropSE. Both proportions and standard errors are multiplied by 100. Fixed definitions and release →
The separately published national estimate is 76.4%, with source standard error 0.2645 pp. It is weighted nationally; averaging state percentages would not reproduce it. National source and definition →
Before comparing two states
- Keep the population with the number. These estimates include internet users aged 15+, rather than everyone living in the state. People who do not use the internet are outside the denominator. This table cannot tell you the participation rate among all residents.
- Read both source errors. Sorting percentages gives an observed order, but sampling uncertainty can affect that order. A pairwise significance test needs a justified variance method; the two marginal SEs alone do not provide full covariance information. Comparing many states also raises a multiple-comparison problem. This guide performs no such tests or population ranking.
- Use one compatible source and period. NTIA's prior-six-month activity question is different from Pew's all-adult platform adoption, HINTS adult visit frequency or a company's active-account measure. No U.S. state comparison here is merged with those measures.
- Respect missing dimensions. These are marginal state estimates. They do not supply state-by-age, state-by-income or platform-by-state intersections. One November 2023 period cannot establish state growth over time.
Source population, questionnaire, method and published cells: NTIA fixed release. Comparison guidance is Social Usage's interpretation of those limits, not a new source-reported ranking or cause.
Follow the source and keep its dates
Question, selection and nonsampling limitations
The activity sequence concerns things done online at least occasionally during the past six months. The social-network question is: “What about using social networks, such as Facebook, Twitter, or Instagram?” Its named platform examples do not give separate platform counts.
One randomly selected internet user aged 15+ per household; skip the activity questions when every internet user is under 15. Proxy responses are allowed. NTIA publishes weighted Census Bureau Current Population Survey supplement aggregates; this guide reads those aggregate cells and processes no individual responses.
Sampling standard errors do not cover every reporting, nonresponse, imputation, proxy-response or other nonsampling error. A smaller SE does not prove that an estimate has no bias. The source cells also do not identify what caused an observed geographic difference.
November 2023 technical documentation, PDF page 148 / Attachment 8, questionnaire page 6 (USEINT / SOCIAL); fieldwork on PDF page 7. Official tabulation code and dictionary.
Reproduce the cells and graphic
- Find
dataset=Nov 2023,variable=socialNetworkUser,universe=adultInternetUserin the source CSV. The retained file uses row 769 including the header; physical rows can change later. - For a state, read its
XXPropandXXPropSEcolumns and multiply each by 100. Keep the source's four-decimal percentage/SE precision in the table. - The illustrative graphic plots California's and Wyoming's source SEs on a 0–3 percentage-point axis. Bar width is SE ÷ 3 × 100. Labels round SE to two decimals only for display. No interval endpoints or margins are derived.
- Retain all 51 state/D.C. inputs, source flags, the original reference period and a separate national benchmark. The full table sorts names alphabetically, without adding rank numbers.
Fixed source release: us-2026-10-01-ntia-v1. Guide record: ntia-state-guide-2026-10-05-v1; count derivation: ntia-2023-state-geography-count. All observation inputs, scope and source-method evidence are bound to the guide's assisted review.
Dates, assisted review and permitted reuse
- Survey fieldwork
- 12–18 November 2023
- Source-table publication
- 6 June 2024
- Retained source retrieval
- 1 October 2026 (UTC)
- Guide checked
- 5 October 2026 · Assisted evidence and uncertainty review
This guide reuses the reviewed immutable source and retained official Explorer code. It is not a new source fetch, a 2026 participation estimate or an independent human editorial review. The original source-release dates remain unchanged.
The selected federal aggregates and original Social Usage presentations have a recorded reuse basis. The linked CSV and fixed participation chart preserve source, scope, period, uncertainty and version. No copied source graphic, agency logo or respondent data is included.
U.S. federal statistical work; 17 U.S.C. § 105 · Sources and rights. Source: National Telecommunications and Information Administration (NTIA), Internet Use Survey, November 2023; U.S. Census Bureau Current Population Survey. Selection and presentation by SocialUsage; no agency endorsement implied.