Everything government, Parliament, the statistical agencies and the major policy institutes published — read, summarised and ranked so you can skim the day in a minute.
Reports public attitudes to climate change and environmental issues from the Opinions and Lifestyle Survey (OPN) for September 2026, presenting social insights on concern levels, perceived risks and support for environmental policies.
Polling by PF Global and IPPR Scotland finds 61% of Scots still back a transition to net zero; 69% want oil and gas decisions made case‑by‑case, 63% back more renewable investment, while 15% and 11% favour removing restrictions or ending drilling immediately.
ONS publishes real-time indicators on economic activity and social change using rapid-response surveys, novel data sources and experimental methods to provide early, experimental analysis of current UK economic and social trends.
Shows Donald Trump’s current approval at about 30% with 68% disapproving; notably, 31% of his 2024 voters now disapprove of his job performance, marking a substantial drop in approval during his second term.
Technical note describing the design, sample, weighting and fieldwork procedures used to run the Small Business Survey 2025, setting out how UK small firms were sampled and surveyed.
Among voters surveyed, 44% say their congressional vote this year is a vote against Donald Trump, while 19% say it is a vote for him, indicating opposition to Trump is a larger motivating factor than support.
Forty-four percent of Americans experienced difficulty affording a basic necessity in the past year, with the burden concentrated among lower-income adults, highlighting widespread cost-of-living pressures.
Since 2019 an increasing share of Americans who say inequality is too high now believe the economic system must be completely rebuilt, reflecting a shift toward more systemic views of inequality and its remedies.
ONS releases Business Insights and Impact on the UK Economy, reporting on how current economic challenges affect firms’ financial performance, workforce, trade and resilience, using recent business response data to track impacts.
Findings from the 2026 inclusion monitoring survey report on digital identity and attribute services certified against UK Government standards, assessing inclusion outcomes for users of these certified services.
Pew’s survey research shows most online communities are not centred on politics, and a majority of members say political discussion should generally be kept out of the communities they visit most.
ONS publishes Health Insight Survey data on patient experiences of accessing NHS services in England. The statistics are funded by NHS England and are presented as official statistics in development, covering access, treatment and service experience measures.
Finds that rising costs of household essentials are the principal economic worry for most Americans, mirroring concerns from four years earlier ahead of the last midterm election.
Reports that 84% of registered voters say the economy is very important to their midterm vote; while the economy tops issues for all voters, Republicans and Democrats differ on the salience of other issues.
About half of US adults report never using chatbots; non-users most commonly cite disinterest, privacy worries and concerns about accuracy as the main reasons for avoiding them.
Most Americans discuss news privately online (via messaging or private groups), while only about a quarter publish news-related posts publicly on social platforms, according to Pew’s analysis of online news engagement behaviours.
Bank of England Research releases the Bank Liabilities Survey for 2026 Q3, a quarterly survey of banks and building societies designed to improve understanding of lenders’ liabilities and capital and their role in credit and monetary conditions.
Bank of England Research publishes the Credit Conditions Survey for 2026 Q3, a quarterly survey of banks and building societies that reports on trends and developments in credit supply, demand and lending conditions.
A Pew-Knight Initiative survey finds about one in three US adults have ever left an online community, based on data from the American Trends Panel conducted in May; the release explores motivations and prevalence of departures.
Describes Wave 198 of Pew’s American Trends Panel: a nationally representative survey conducted 24–30 August 2026, with 5,166 responses from 5,602 sampled panelists, and details on weighting and fieldwork.
Global public‑opinion polling finds widespread concern that social media harms democracy by spreading rumours; accompanying survey highlights anxieties about AI, job losses, and inequality in some countries.
Tables report confirmed toxigenic Corynebacterium diphtheriae cases among asylum seekers in England covering 2022–2026, detailing incidence by year to support public health surveillance and outbreak response planning.
Findings from the 2024/25 Northern Ireland Safe Community Survey on experience of crime, attitudes to crime, policing and justice, and perceptions of organised crime, reporting public sentiment and victimisation patterns.
About 38–39% of US adults belong to an online community such as Facebook groups, gaming communities or Reddit fandom forums, which function as virtual ‘third spaces’ for social connection and shared interest.
In a multinational poll, 85% of French adults say their national economy is doing badly — the highest share among 18 high‑income countries included in the survey.
Publishes the September 2026 Decision Maker Panel, a monthly survey of chief financial officers across small, medium and large UK firms used to monitor economic developments and businesses’ views.
Public opinion polling summarises US religious groups’ midterm voting preferences, noting differences between denominations — for example, Catholics leaning three points towards Democratic House candidates — based on Pew’s recent analyses.
Survey experiments using Opus 4.6 and GPT‑5.1 produced markedly different synthetic public‑opinion results, with neither model’s output matching actual human survey benchmarks, highlighting model-dependent distortions in AI polling.
US polling finds the Iran conflict is widely unpopular: 65% disapprove of President Trump’s handling of US military action, 33% approve, and many Americans expect the conflict to continue for months.
Digital twin synthetic surveys performed poorly at estimating public sentiment overall and were especially error‑prone for some subgroups, notably Republicans and Black adults, calling into question AI stand‑ins for human survey‑takers.