The healthy life expectancy is the average number of years that one can statistically expect to leave before falling sick because of some serious, chronical disease. Professor Ugo Bardi, teacher of Physical Chemistry at the University of Florence, recently noted an impressive decrease, starting in 2003, in the healthy life expectancy of Italians and therefore asked: how long do Italians really live?

Personally, I have no specific competence to answer this question, or to validate the correctness of Bardi's starting thesis. If that's what you were looking for, read directly that article and other posts on the same topic by Bardi. The only reason I am talking about that post here is because it provides a perfect example of a general problem, which is important for all citizens, on which I'm working these days.

Bardi may be absolutely right, or he could also be presenting, for whatever reason, a completely wrong picture. As far as I am concerned here, both hypotheses have the same probability, and I'm listing both of them just to make the central point I want to make here clearer. In a footnote to the article, Bardi writes that:

"Is it true that life expectancy at birth in Italy continues to increase? ... [I] do not know who to believe between CIA and ISTAT (the Italian Institute for Statistics); however, I noticed a worrying thing in the ISTAT tables: data are provided as estimates, without further explanation. From that I conclude that they are extrapolations, not experimental data. In fact, EUROSTAT leaves white boxes for data on life expectancy in Italy from 2008 to 2010 ... What has happened that made data about Italy disappear?"

If it's true that Italians live less than they did ten years ago and get sick earlier, this is a really serious to which we should react now. If it's not true, that's great, but then it should be evident to everybody and backed by official data.

And this is just the thing that I want to stress: Pardi might be right or wrong, but why can't he, or any other expert with the right skills to assess the valididty of Pardi's graph, find all relevant, up to date official data online?

ISTAT is (quoting from its official website) "a governmental research organization ... the main producer of official statistics in support of [Italian] citizens and policy makers."*

The data we are talking about are public data, consisting of simple numbers and dates that can be published without raising any privacy issue. ISTAT is a public institution, paid by taxpayers to protect their interests. Besides, today there are computers and the Internet, which allow you to publish whatever you want online at the lowest possible cost, perhaps automatically. So why should there be this uncertainty? If those data had been already collected, then they must be published online. If they have not been collected yet, perhaps simply because of lacks of funds, then, at least, the reason should be clearly stated online and in all the summary tables, right below the notes that say that certain numbers are just estimates.

This increase in life expectancy is only one case, even if it is an important one. This is a very general issue. I've already provided other examples on this website, from the walkability of cities to water management and road maintenance. Another examples of public data that are much more obscure than they should is in the Eurostat energy data: "Energy is an active are of EU public policy. Yet authorities are not revealing information... that is crucial to determine whether its policies are distorting the market and come at too high a cost to society."

The solution to problems like the uncertainty on life expectancy is always the same: unless there are real privacy issues, public data created or used by Public Administrations to serve citizens should always be published online, in order to increase transparency in politics, save public money and stimulate the economy. Unpublished data should become exceptions, to justify every time providing valid reasons, publicly. I have discussed this and the above reason for openness in the Open Data, Open Society report.