Methodology
MMT Pulse counts how Modern Monetary Theory and its ideas appear in UK public life, week by week. It shows correlation, not causation: it can show that something changed and when, not why.
What counts
An item counts when it does one of three things: names MMT; uses an MMT-consistent idea without the label; or uses an opposing frame. The ideas and frames are concepts in the Knowledge Graph, each with its own search vocabulary:
- MMT by name (mmt by name)
- Currency issuer (mmt-consistent idea)
- Taxes don't fund spending (mmt-consistent idea)
- Spend first, tax later (mmt-consistent idea)
- Sectoral balances (mmt-consistent idea)
- Deficit = private surplus (mmt-consistent idea)
- Job guarantee (mmt-consistent idea)
- Functional finance (mmt-consistent idea)
- Real resource limit (mmt-consistent idea)
- Deficit myth (mmt-consistent idea)
- Taxpayers' money (opposing frame)
- How will you pay for it? (opposing frame)
- Maxed-out credit card (opposing frame)
- Fiscal black hole (opposing frame)
- Household budget (opposing frame)
- Magic money tree (opposing frame)
"MMT" on its own also means a fuel additive, a telescope and more, so it only counts beside an economics word, and known other meanings are excluded. Items that fail these checks are kept out of every figure.
How framing is judged
Each item is classified by a language model into one of these stances:
- Endorses MMT
- Neutral mention
- Hostile to MMT
- Misrepresents MMT
- MMT-consistent idea, unlabelled
- Uses an opposing frame
The headline ratio compares Endorses MMT and MMT-consistent idea, unlabelled against Uses an opposing frame and Misrepresents MMT. Neutral and hostile mentions of MMT are reported beside it, never inside it. Items from the last 8 weeks are all classified; older weeks are classified on a random sample of up to 150 items per source per week, weighted back to each source's total, and shown with a 95% interval.
Not yet published Framing figures stay off this dashboard until the classifier has been measured against at least 300 hand-labelled items and reaches a macro-F1 of 0.7 and a Cohen's kappa of 0.6. Until then, pages show volume only.
Sources
OpenAlex (academic citations)
Scholarly works that cite a list of core MMT works, or discuss Modern Monetary Theory in their title or abstract, from OpenAlex (CC0). Academic metadata arrives weeks or months after publication, so the latest weeks are always incomplete. Compared with its own trailing average rather than all economics publishing, which is too large to fetch within the free allowance.
Google Trends
Google offers no open API for Trends: its official API is a closed alpha and unofficial scrapers are blocked. Search interest is shown only when staff upload a Trends export by hand, and is marked with the date of the upload.
Wikipedia pageviews
Daily views of English Wikipedia articles on MMT and its core ideas, from the Wikimedia Pageviews API (human traffic only, bots excluded). Shown beside a basket of comparator articles (Keynesian economics, monetarism, austerity, fiscal policy, government debt) so a rise can be read against general interest in economics. Pageviews are global and cannot be split by country.
The Guardian
Full-text search of The Guardian's archive through the Guardian Open Platform, with weekly totals from the politics, business and UK news sections as the baseline. One outlet, so it shows how the debate travels in one national newspaper, not the whole press. Content courtesy of the Guardian; excerpts link back to the original.
UK news outlets (Media Cloud)
Daily counts of stories using the dashboard's phrases from a fixed panel of UK outlets that Media Cloud has covered since 2021: BBC (English), The Independent, Daily Mirror, Evening Standard, Daily Express. Counted, not read: Media Cloud cannot share article text, so these stories add to volume and are not classified for framing. The Guardian is counted through its own API instead. Not covered: the Daily Mail, The Times, The Sun, the Financial Times, Sky News, The Spectator and the New Statesman. The Telegraph is only partly covered, and GB News and the i only recently, so they are left out. The baseline is every English-language story the panel published each day.
Movement activity (Sovereign Economics tools)
Organisation-level counts from the Sovereign Economics tools: public events listed on the events calendar, social posts published, publications from member organisations, and views of embedded knowledge widgets. Counts organisations' activity, never individuals.
UK Parliament Hansard
Spoken contributions in the House of Commons and House of Lords, from the UK Parliament Hansard API. The baseline is the total number of spoken contributions in the same week, so a busy week is not mistaken for growth. Hansard is published with a lag of a few days. Contains Parliamentary information licensed under the Open Parliament Licence.
Bluesky
Public Bluesky posts in English matching the dashboard's search terms, with likes and reposts as reach. Bluesky has no location data, so this is English-language conversation, not UK-only. Individual posters are never named or ranked on this dashboard.
Not available. Reddit announced on 30 September 2026 that it is closing its public Data API: no new access after 31 October 2026, RSS feeds end on 13 November 2026 and remaining public access ends in March 2027. After that, access needs a commercial agreement with Reddit, so the dashboard does not collect from Reddit.
YouTube
Videos found by the YouTube Data API for the dashboard's search terms, with view counts as reach. Searches are biased to UK viewers but channel country is often not stated, so this is English-language conversation rather than strictly UK. The API's daily quota limits how many terms can be searched.
Limits
- Volumes depend on what each source indexes. A change in a source's coverage can look like a change in the debate.
- "UK news" is the Guardian plus daily counts from a fixed panel of UK outlets through Media Cloud (BBC (English), The Independent, Daily Mirror, Evening Standard, Daily Express). Panel stories are counted, not read, so they add to volume but not to framing. It is not the whole press. Not covered: the Daily Mail, The Times, The Sun, the Financial Times, Sky News, The Spectator and the New Statesman. The Telegraph is only partly covered, so it is left out.
- Social figures are English-language, not UK-only: Bluesky has no location data and YouTube channel country is often missing.
- Academic metadata arrives late, so recent weeks undercount.
- Individuals who are not public figures are never named or ranked.