It looks like I spoke too soon in calling John Cox’s Neighborhood Legislature initiative dead. And I could still be proven wrong in taking Cox to task for having the initiative fail just short of the number of signatures.

Challenges to the full count that found the initiative coming up just short continue. And there are reasons to wonder about the reliability of the count—and the system itself.

I’ve talked to more people since I wrote about the initiative’s narrow failure a week ago, and I was too hard on Cox. He did many of the right things on the initiative—he spent years developing it and refining it, with far more care than many initiative proponents. And when it came to qualification, he started early, in an apparent effort to save money and make sure he qualified.

And it appeared the initiative had enough to qualify on a spot check of signatures and avoid the full check.

But counties behaved in ways that raise questions about the check.

Suspicion is in order here not because Cox is a candidate for governor (as a Republican, he is thought to have little shot of winning, despite his rise in the polls), but because the idea of a larger legislature is dangerous to powerful interests. And we’ve seen interests in recent decades put their foot on the scale of the initiative process, changing timing (all initiatives in November) and titles and summary (by pressuring the attorney general’s office). The Neighborhood Legislature was targeted on the title and summary as well (which is why we should have citizens’ committees write titles – and initiative language, but that’s another, longer story).

The trouble is that Cox and Mike Arno’s petition company were playing in a different initiative world that makes it hard to save money. Signature gathering has become harder and more expensive. That’s in part because the places where you can circulate petitions are limited (grocery stores are the latest to prevent petition circulators from operating outside their doors). Part of the problem is that people do more shopping at home online.

Many signature gatherers are retired, and the younger core has gotten used to making more money, and is not quite so eager to work. (A shocking number has turned to bitcoin related enterprises, so maybe they’ll be back to work more when bitcoin melts down).

The Democrats’ decision to change initiative timing – to put all measures on the November ballot – has meant that there are more measures competing for signatures on the same time line.

With all these changes, Cox and Arno miscalculated in not having a bigger cushion. But they may have had the signatures anyway.

The problem is we don’t really know, because there are real questions about the full count conducted by the counties.

One problem that we’ve seen with ballots – where those who voted have their ballots not counted because their signature has changed from the one on their registration card – seems to be rearing its head here, with high numbers of signatures disallowed for lack of signature match.

Also, are counties – with understaffed clerk’s offices—really checking every single signature? It’s suspicious that most of the signature checks seem to have occurred right on the deadline day. Worse still, some counties are offering less than full cooperation in re-examination of their full count. Santa Clara County has issued restrictions that prevents those examining the county’s count from having water to drink (essential for human health) or electronics (iPads or laptops are vital to be able to check different databases when you’re checking the checkers) in the county offices when they examine signatures.

It’s high time for an independent audit of how the counties do this.

Indeed, as I’ve learned more about the questions around the handling of Cox’s measure, I wonder about the entire initiative system. It’s more dependent than ever on money. Pieces of the process – signature gathering, signature checking — seem withered, even broken. Now would be a good time for a review and a reimagining.