Analysis · No. 6

Did they listen?

What happens after a UK government consultation: how often the plan changes, how long the answer takes, and how many answers never come.

Seam analysis by Matt Bevington (how to cite) · first published 9 October 2026, refreshed monthly · Data to 1 November 2026 · How we did this

2,191consultations closed since October 2021
77%have a published response
174 daystypical wait for a response
24%answered within the 12 weeks government aims for
38%of responses read say the plan changed

What government decided

We read 529 published responses. The most common outcome was “proceeding with changes”: 38% of them, or about 4 in 10.

Otherwise, 18% said it would go ahead as proposed, 5% that it wanted more consultation or a review first and 2% that it had dropped or paused the idea.

For 37% we could not tell. That is 194 responses. In these, the summary and any web-page version did not say, or no quote from the text backed the reading. Leave those out and 61% of the 335 clear answers involved changes.

What the response said would happen

529 GOV.UK consultation responses, read one by one. Share of all responses read.

  • Proceeding as proposed18%
  • Proceeding with changes38%
  • Not proceeding or paused2%
  • Further consultation or review5%
  • Unclear37%

“With changes” means the response itself says something was changed, dropped or added. It does not say whose comments led to the change.

What each label means
LabelWhat it meansResponses
Proceeding as proposedThe government says it will go ahead with what it consulted on, without material changes.96
Proceeding with changesIt will go ahead, but has changed, dropped or added something compared with the proposal.203
Not proceeding or pausedIt has dropped the proposal, or put it on hold.10
Further consultation or reviewNo final decision yet: it plans more consultation, a review, a pilot or more evidence first.26
UnclearThe summary and any HTML response do not say, or the quote offered could not be found in the text.194

Has it changed over time?

The same results, by the year the consultation closed. We leave out years with fewer than 20 responses read.

Outcome by year the consultation closed

Share of responses read in each year. Recent years are incomplete because their slow responses have not arrived yet.

  • Proceeding as proposed
  • Proceeding with changes
  • Not proceeding or paused
  • Further consultation or review
  • Unclear
  • 2021 (23 read)17%30%13%39%
  • 2022 (125 read)14%44%37%
  • 2023 (134 read)20%37%31%
  • 2024 (105 read)21%34%37%
  • 2025 (111 read)14%41%42%
  • 2026 (31 read)35%29%35%
Show the numbers
Closed inReadProceeding as proposedProceeding with changesNot proceeding or pausedFurther consultation or reviewUnclear
20212317%30%0%13%39%
202212514%44%2%3%37%
202313420%37%4%7%31%
202410521%34%2%6%37%
202511114%41%0%3%42%
20263135%29%0%0%35%

How long the answer takes

The government’s own consultation principles say a response should be published within 12 weeks of a consultation closing, or the department should explain why not.

Of the 2,036 consultations that closed at least 12 weeks ago, 495 had a response within that time: 24%, or about 1 in 4. The typical wait was 174 days, about 25 weeks.

Time from closing date to published response

1,633 consultations with a response. Hover a bar for the count.

  • Within 12 weeks508
  • 12 weeks to 6 months555
  • 6 months to a year352
  • One to two years180
  • More than two years38

The response date is the day GOV.UK marked the page as answered. A page edited again later can carry a later date than the real one, so a few waits will be overstated.

By the year the consultation closed

The typical wait includes consultations still waiting, for as long as they have waited. So quick answers alone do not flatter recent years.

Closed inConsultationsAnswered so farWithin 12 weeksTypical wait
20219093%24%179 days
202245689%27%174 days
202345888%28%153 days
202438077%22%223 days
202541681%22%168 days
202639139%20%194 days

2021 starts in October, five years before the data was taken. The latest year is incomplete.

The ones that never came

258 of the 1,743 consultations that closed more than a year ago still have no published response: 15%, or about 1 in 7.

Some may still come. The chart shows how many are still waiting at each age.

Still waiting, by when the consultation closed

Share of consultations with no published response yet.

  • Closed 1 to 2 years ago73 of 41018%
  • Closed 2 to 3 years ago87 of 38922%
  • Closed more than 3 years ago98 of 94410%

By department

Among the main departments, Education answered 33% of its consultations within 12 weeks, and HM Treasury 7%. A department or body with fewer than 20 cases in a column shows “too few to say”.

Responses by department

The main Whitehall departments, and any other public body with at least 20 consultations.

Department or bodyConsultationsAnswered within 12 weeksTypical waitNo response after a yearResponses readWith changesAs proposedNot proceeding
Housing, Communities and Local Government17315%179 days16%3658%11%0%
Energy Security and Net Zero17016%196 days11%3944%3%0%
Environment, Food and Rural Affairs16719%235 days14%3247%19%0%
Business, trade and science (DBT, DSIT, BEIS)16323%165 days15%3537%11%3%
Transport12217%274 days12%3247%19%9%
HM Revenue and Customs12119%164 days24%18too few to saytoo few to saytoo few to say
Competition and Markets Authority11444%88 days1%449%57%5%
Education10433%134 days14%14too few to saytoo few to saytoo few to say
Health and Social Care10030%181 days14%2744%19%0%
Natural England9967%41 days16%284%4%0%
Home Office9623%219 days14%3749%11%0%
HM Treasury917%197 days14%2339%17%4%
Ministry of Justice6514%241 days19%18too few to saytoo few to saytoo few to say
Environment Agency6430%149 days10%2450%21%0%
Building Digital UK484%301 days4%8too few to saytoo few to saytoo few to say
Maritime and Coastguard Agency4651%84 days5%15too few to saytoo few to saytoo few to say
Culture, Media and Sport4526%163 days3%15too few to saytoo few to saytoo few to say
Work and Pensions4329%120 days10%5too few to saytoo few to saytoo few to say
Office of Rail and Road4138%145 days3%18too few to saytoo few to saytoo few to say
Marine Management Organisation396%not yet half84%0too few to saytoo few to saytoo few to say
Ofqual2857%71 days0%11too few to saytoo few to saytoo few to say
Medicines and Healthcare products Regulatory Agency274%372 days33%4too few to saytoo few to saytoo few to say
Ministry of Defence16too few to saytoo few to saytoo few to say1too few to saytoo few to saytoo few to say
Cabinet Office15too few to saytoo few to saytoo few to say6too few to saytoo few to saytoo few to say

Some examples

We pick one of each kind by rule: the most recent response with that label whose short quote states the decision. Quotes are word for word from the response.

If you respond to a consultation

  • Expect a wait. The typical response took 174 days, and 24% came within the 12 weeks government aims for.
  • Some never come. About 1 in 7 consultations that closed over a year ago have no response.
  • Read the response, not the headline. 38% of the responses we read said the plan had changed. The response itself sets out what changed.

See every open consultation, and when a response is likely, with Ask.

How we did this

Method and caveats

Data. Every consultation on GOV.UK that closed between 1 October 2021 and 1 November 2026: 2,191 in all, from the GOV.UK Search API. A consultation counts as answered when GOV.UK lists it as a consultation outcome. We do not include consultations run only on a department’s own website, or by the devolved governments.

Timing. The wait is the number of days from the closing date to the date GOV.UK marked the page as answered. 46 responses dated before their own closing date are left out of the timing figures. Usually the closing date was moved after the fact. The typical wait is a median that counts consultations still waiting, for as long as they have waited (a Kaplan-Meier estimate). So quick answers alone do not flatter recent years. The 12-week share counts only consultations that closed at least 12 weeks ago.

What was decided. For 529 answered consultations we took the outcome summary GOV.UK publishes on the consultation page. Where the response is also a GOV.UK web page, we took the text of that page too. We asked a language model (claude-haiku-4-5) to choose one of five labels using that text alone, and to copy out a sentence that supports it. A computer check then looks for that sentence in the text, word for word. If it is not there, we mark the case unclear. That happened 132 times. 295 of the responses read had a web-page version. 117 were published only as PDFs or other files, so only their short summary was read. The other 117 had no response document attached, so the outcome summary was all there was to read.

The labels. Proceeding as proposed: the government says it will go ahead with what it consulted on, without material changes. Proceeding with changes: it will go ahead, but has changed, dropped or added something compared with the proposal. Not proceeding or paused: it has dropped the proposal, or put it on hold. Further consultation or review: no final decision yet: it plans more consultation, a review, a pilot or more evidence first. Unclear: the summary and any HTML response do not say, or the quote offered could not be found in the text.

Limits. “Proceeding with changes” says the plan changed, not why, and not whose comments caused it. A department may change a plan for reasons that have nothing to do with the consultation. PDF responses are not read, and a short summary often says only that a response has been published, which is why many of those cases are unclear. A model can misread a long response. The quote check stops it inventing support, but does not catch every misreading. The worked examples are picked by a fixed rule, so you can check them yourself.

Groups. Renamed departments are counted with the department they became (BEIS with DBT and DSIT, DLUHC with MHCLG). Any group with fewer than 20 cases shows “too few to say” rather than a share.

← All analysis