If we are to understand why TfL assumes some projects work better than others, we need to see the workings
I asked at the London Cycling Campaign yesterday exactly what TFL made known about cycling and other traffic assumptions, when they calculate changes that result from new cycling infrastructure. People at the meeting thought that the journey times and traffic density calculations came from software that is probably using proprietary algorithms, and today on twitter, I was told that even the raw data has not been released.
Why would we want this data? Cyclists observe that shifting people onto bikes reduces pressure on public transport and private cars, and is likely to help reduce vehicle traffic. It is unclear how much TfL accept this argument when calculating road use changes.
London has claimed to be at the forefront of Open Data policy under Boris Johnson’s administration. Many people have been impressed, or at least pleased with the availability of data, for instance through London Data Store. It is on the face of it surprising that raw traffic data might not be available.
I could however only find some very general statistics about cycling levels. Additionally, the site’s function to request new data sets was “temporarily suspended” because of spam in late 2012. It is still switched off. The site advises:
If you're looking for a specific piece of information rather than a raw figures [sic], you may be better off making a request under the Freedom of Information Act.
Either way I wondered what other people thought about making a concerted effort to persuade TfL and the Mayoralty to release any raw data that would be helpful, and also to propose what data is needed to start analysing the likely effect of possible changes. If in fact some of this data is available, it would be great to know that too.
If the final traffic projections are indeed produced through proprietary systems, that will be trickier to obtain. It would raise serious questions about the accountability of government decisions, if they are based on unaccountable and non-examinable algorithms.
On Twitter, Pascal Van den Noort pointed me to similar data released in the Netherlands. In any case I am not the person to start analysing such data, I am merely someone who would advocate that an informed debate starts with an informed public, and that in today’s world means that we all get to see the data and workings so we understand why our institutions are making the arguments that they do.