about my county st johns county florida

At the February 6, 2024 Board of Commissioners meeting, Ann Taylor made a public comment, suggesting that the minutes of Board Meetings should include pertinent information about each agenda item that is voted on.

Ms. Taylor described the difficulty in conducting an analysis of Commissioner voting records.  I suspect she was recounting the challenge of coming up with the voting records of Commissioners Whitehurst and Dean that have been splashed all over social media and blast emails.

I completely agree with her. But I’m not sure that her statement is helping the cause they have been communicating about commissioner voting records.

I agree with her because I tried to do the same thing and learned the hard way that the way the County Board minutes are taken and posted is highly narrative. If you want to follow authentic research protocol, it is very difficult to arrive at any conclusion that could not be easily refuted.

As Mark Twain is believed to have said, “there are lies, damn lies and statistics.”  Numbers are only as good as the context within which they are presented.  With a good spin, you can make any point you need to make.

I know. Years ago, I analyzed salary data for top bank executives. I presented my analysis using the method I’d been taught.  It showed the CFO was overpaid.  Oops, I was told. You forgot to give credit for x,y,z, I was told.  Add 25% here, 30% there.  All of a sudden, the CFO deserved a healthy raise.  The spin was that his job was “bigger” than what was described. It was one of my first lessons in “the way the world works.”  If you want to read more about my opinion of executive pay, look here.

Here is the problem with analyzing Commissioner voting records

Because the minutes are narrative – words rather than numbers – it is not possible to complete a quantitative analysis of that data without a great deal of pre-work. To do that you have to turn the words into numbers.  You have to categorize each word, agree on the categorization, and use that category consistently throughout the analysis.  Only then do you count.

As an example, the voting record that is being presented in social media represents “new” development.  Well, that is interesting because “new” is not a descriptor in the minutes.  To get a good data point, one must take each of the agenda item codes and categorize them as “new.”

Given the history of development in the county, that’s a difficult data point to define.  Does “new” mean never before the Board?  If so, how does that work with development that was approved years ago but is only now being developed?

Ms. Taylor’s comments about including pertinent information are right on target.  By creating the categories up front, and then collecting data in agreed upon categories, you can achieve a credible analysis.

But that is not what we have here.  Whoever did the analysis and arrived at the voting records that have been splashed all over social media made their own decisions about categories so that they could turn a narrative into a quantitative analysis that fit their narrative – the commissioners approved all development.  Are they correct? 

Numbers are enticing. I saw a comment on social media that said something like, “now we have the data.”  Sure, you have data.  But is it credible data? 

Or do you just want so badly to believe the commissioners are in the pocket of developers that you’ll believe it without question?

I’ll close with one of my favorite quotes:  “Most people use statistics like a drunk man uses a lamppost; more for support than illumination”  ― Andrew Lang