We conducted an analysis of responses to MHCLG’s (the Ministry of Housing,
Communities & Local Government) Planning for the Future White Paper consultation. The consultation invited stakeholders to submit their views on reforms for plan-making, development management, development contributions, and other policy-related proposals aimed at modernising the planning process.

MHCLG received over 50,000 responses, of which 10,000 included written answers to open-ended questions submitted through an online platform as well as unstructured email responses which required comprehensive thematic analysis.

Within a timeframe of 2 months, we thoroughly processed all responses to produce a summary report. The report identified the most important emerging themes for each question, broken down by response group and other relevant aspects.

Our team of researchers and data scientists designed a methodology involving a combination of manual reading and software-assisted processing of responses to allow the delivery of our analysis on a tight timeline, while ensuring all views were taken into account.

The general framework of our approach involved machine-assisted human coding. In this setup, human coders go through the text, carefully reading a large sample of consultation responses and using qualitative research techniques to produce a codebook. On the basis of this, machine learning is then employed to code the remaining responses.

This approach maximised our ability to draw out a mix of qualitative insights (e.g. interesting ideas proposed by a small number of respondents) and quantitative insights (e.g. establishing which views were most widely held among respondents).