How to cite SHIFT

SHIFT measures how advertised demand for work changes over time, by kind of work and by market. It exists to be quoted — and this page says exactly how, so that nobody has to guess.

What may be quoted, and what may not

In plain terms:

• Yes — any figure published on this website, quoted with attribution, in reporting, research, teaching or commentary.
• No — republishing the paid occupation documents, in whole or in part, or reproducing this website.
• No — presenting a SHIFT figure without saying where it came from, or restating it in a way the method does not support. If in doubt, the method is public and we will show how a particular figure was produced.

The wording to attribute it with

In running text:

A link to the page the figure came from is welcome and is not required.

Read this before quoting a figure

• Methodology — what is counted, what is left out, and why. Every rule is fixed before the counting begins. Read this before quoting any figure.
• About — who publishes the index, how it is funded, and the rights under which the underlying job-advertisement information is used.
• Kinds of work — the index by kind of work. Each page states the direction of change and the searches behind it.
• Questions and answers — the questions people actually ask, including what the figures are not.
• Sample report — a full occupation document, so the paid product can be inspected before it is bought.

The open dataset, and how to cite it

The occupation margin of Edition 1 is published as an open dataset under CC BY 4.0, with a permanent identifier, so a figure quoted from it can be traced to a fixed version.

10.5281/zenodo.22346056 names this version; 10.5281/zenodo.22346055 always resolves to the most recent one. The same files, with the generator that writes them, are at github.com/Shiftindex-org/shift-labour-demand-index.

It carries the occupation margin only — one row per kind of work. It does not carry the occupation-by-market detail, which is what the paid reports are.

The same open dataset is mirrored on Hugging Face (JoseCovas/shift-labour-demand-index) and Kaggle (josecovas/shift-labour-demand-index-2024-2026-edition-1), and is indexed by OpenAIRE.

The methodology, as a citable paper

The measure, its aggregation rule and the pre-registered robustness tests are set out in a methodology paper (a data descriptor), published open access with its own permanent identifier.

10.5281/zenodo.23133070 names this version; 10.5281/zenodo.23133069 always resolves to the most recent one. The paper describes the open dataset above and contains only the public methodology.

What is not published, and will not be

SHIFT publishes derived aggregate counts and indices. It does not publish, and does not licence, the underlying job advertisements, the provider datasets it draws on, the full list of search terms, or the collection code. That is a limit of what is offered, not an invitation to ask.

Public statistics and third-party benchmarks are used only as external context, are identified where used, are not presented as SHIFT data, and remain subject to the rights and terms of their owners.

If you are writing about this

Press and general enquiries: contact@shiftindex.org. We are available to demonstrate how a particular figure was produced, and we would rather answer a question before publication than correct a sentence after it.

SHIFT is published in an independent capacity and is not affiliated with, endorsed by, or representative of the views of any standards body, professional institution, trade association or data provider. Nobody pays to appear in the index, and nobody can pay to be left out of it.