, .
France and Britain both have a national industrial champion in jet engine design and production. Britain has Rolls-Royce. France has Safran. At the moment Safran is doing better, but things can change quickly in that industry.
France and Britain both have a national strength in government open data and regularly top international rankings for the quality and breadth of publicly available government data that is free for anyone to download and use without restriction.
How well do these two national strengths overlap?
Britain's Companies House website and bulk data downloads tell us about Rolls-Royce Plc. We know that it is registered at Kings Place, 90 York Way, London and is active in the field of 70100 - Activities of head offices. The company mostly operates in Derby, but that's not where it's registered.
France's Sirene website and bulk data downloads tell us about Safran Nacelles. The website is much worse than Britain's Companies House website so I can't link to the listing.
We know that Safran is registered at 78117, Chateaufort where it does 30.30Z - air and space construction and employs 200 to 249 people. It has further operations at 27500 Pont-Audemer, where 50 to 99 people do 33.16Z - repairing jet engines and at 31770 Colomiers (250 to 499 employees), 57190 Florange (100 to 199 employees), and 76700 Gonfreville L'Orcher (1000 to 1999 employees) where the employees do 30.30Z - air and space construction.
There is much more information about both companies in further places. I am happy to sell you some of it via The Data City where I am a founder, a director and the CTO. But I am determined to do the hard work and keep this blog post simple. You don't need to know here that the French data includes Safran's operating locations in Burnley, UK or Casablanca, Morocco for example. So you'll have to trust me that the above simplification is a reasonably fair one. Or stop reading if you don't.
What does the French and British data look like on a map?
The French view of Safran's operations via Sirene includes multiple locations, what occurs at each location, and the approximate distribution of employees. It is far more detailed than the British view of Rolls-Royce's operations via Companies House. You may be thinking that the UK government has all of this data internally and puts it to good use. Let me assure you that it does not. Closed data within the UK government is rarely used anywhere near as well as open data.
I think that the French data is better. I am not certain that this matters, though I think that it does. There is a reasonable argument that the UK government and broader British society doesn't need to know where British businesses operate since it should overwhelmingly be leaving business activity to the market. A French government, this argument would go, needs to know more about its industrial base because it is so much more interventionist.
Since I think that the idea of a non-interventionist British government is largely a myth I find this argument unconvincing. I think that the UK government has always done interventionist industrial strategy and as it continues to do so it needs better open data to do it better.
Where Britain does better is in the quality of the websites that deliver companies data.
The British site loads faster, has a more modern and slick UI that is consistent with other UK government sites, is more accessible, and works better on mobile. We can measure this with Google's lighthouse tool.
This is just one example of something I've lived and worked through for the past decade when it comes to state data. Both the UK and France have done well, but in importantly different ways.
The UK has focused on user experience and has won. I don't think anyone reasonable would deny that the British government's website is better or that this superiority extends more widely across digital government in the two countries.
The French meanwhile have focused on data. We understand much more deeply how and where Safran operates from the French data than we understand how and where Rolls-Royce operates from the British data. This experience is consistent across this database. The pattern is repeated consistently across many further datasets shared by the two governments.
There are of course exceptions and considerable variations. But overall this difference in national priorities has been consistent. My preference is the French approach to data, but I suspect that each country is getting what most people within that country want. The British preference does seem to be for a better and more consistent user experience and a focus on accessibility over better data. My favourite example of this is the widespread celebration of the UK government's single voter registration website, despite the voter registration rate in the UK being below that in France.
Of course we could in theory have the best of both worlds. We could have both better data and a better website. But I'm much more sceptical of "why not both?" arguments than I was a decade ago.
Both the French and British governments have limited resources to apply to their national data efforts. And importantly the two goals of data quality and website quality are partially in conflict — the UK government website guidelines that make the user experience better are much stricter and slow down the development of new features or the presentation of new data to users.
And that's how I arrive at what I'll no doubt be told is an absurd thought. That maybe if the UK government wants to do industrial strategy with data better, or other things with data better, it should think about getting some obstacles to that out of the way, even if that means letting its own websites get worse. If you think that's a terrible idea, I suspect you're British. If you quite like it, maybe you're French.