After the November election, a number of editorial articles came out with variations on the following theme – that the various depressing voter I.D. laws didn’t actually have much effect on elections. Because hey, the Republicans knocked the elections out of the park, right? That happened despite disproportionately good turnout among minority voters in some elections.

Such editorials are music to the ears of the extreme right (“Yeah. See, we’re not racists. If we were racists, we would have had to turn the firehoses on the swarms of minority voters that voted in the November 2014 federal elections. But we didn’t have to use firehoses or police dogs, or even fire any tear gas into the polling places. Therefore, we have conclusively demonstrated that we’re not racists. Plus, we love democracy. Democracy was very good to us.”) and to the defense attorneys representing the State of Texas in its appeal of Veasey v. Perry et al.

I mean, if an editorial by Nate Cohn in the New York Times, (that notorious hotbed of centrist legal opinions), comes out saying that voter I.D. weren’t no big deal, then the centerpiece of the plaintiffs’ evidence against the State in Veasey is undermined, and the State should walk on all charges, right?

Mr. Cohn was extending a dialog he had begun with the Brennan Center on this question with an earlier and similarly dumb editorial along the same themes.

Mr. Cohn’s argument runs like this: It must be the case that 600,000 Texas voters weren’t disenfranchised because if they had been disenfranchised, they would have been turned away from the polls. Since 600,000 people weren’t turned away from the polls, 600,000 people weren’t disenfranchised. Quod erat demonstram. Ipso facto. Res ipse loquitor. Et cetera.

And because only a fraction of all registered voters lack sufficient I.D., (say, 9 percent, or 11 percent, or something like that) it means that voter I.D. laws don’t actually change election outcomes. Right? Right? Because if the bad guy won by 20 percent, then even if all the voters without i.d. had been able to vote (and even if all of them had voted for the bad guy’s opponent), the bad guy would still have won by 11 percent, or 9 percent. Or whatever.

I’ve written about this before, but maybe if I write slower and louder, people will begin to understand.

The secret to getting people not to vote is to discourage them from attempting to vote. And because the cost of voting is so high compared to the benefit, it is very, very easy to discourage people from voting.

And when people don’t vote, they don’t make an ostentatious show of not voting. They don’t gather en masse, carrying placards and linked arm-in-arm at the polling place to not cast their ballot, in a vast silent repudiation of their civic authority. Such mass demonstrations by non-voters would be quite satisfying, if they happened, and would make for some fun political theater. But they don’t happen, because showing up at a polling place to not vote is almost as much bother as showing up at a polling place to vote.

Non-voters don’t go out of their way to proclaim their nonvoting status. They go to work or school. They run errands. They watch T.V. Some of them register to vote, but fail to register on Mr. Cohn’s radar, and so he assumes that they don’t exist. Some of them fail to register to vote, period, and so they don’t even show up in the statistics of “registered voters who lack i.d.”

The authors of the sociological studies at the heart of the plaintiffs’ case in Veasey v. Perry didn’t merely rely on the self-selecting sample of voters who actually go to vote. They called people up and asked them direct questions about their status as voters, and about their inclination to vote or not vote, based on their circumstances.

As it turns out, (as amply demonstrated by our most recent election) disenfranchisement works. It would have been embarrassing to the arch-conservative proponents and apologists for picture i.d. requirements  if disenfranchisement hadn’t worked, because it’s actually kind of an administrative pain in the ass to exclude those naughty liberals from the polls, and it costs money to administer draconian voter i.d. laws. Luckily for the right, everything went swimmingly.

Let’s look at the November 2014 elections in one Texas county, and see how disenfranchisement succeeded in that county.

Harris County, Texas

Voters in Harris County had a long ballot to work on this year, in part because this very heavily populated county has an enormous county and district judiciary to handle the legal caseload generated by its roughly 4,330,000 residents. As it seems to happen every four years, the district and county courts were filled by Republican candidates, who pretty uniformly won their seats by consistent margins, just as they had even when Bill White carried Harris County in 2010.

2010 was a watershed year for the county – it was the first Federal decennial census in which Harris County had more than 4 million people. The City of Houston, meanwhile, was estimated to have just a smidge under 3 million people; there’s no doubt that Houston, like much of urban Texas, is experiencing a tremendous population boom. In fact, the county has added roughly 240,000 people or more since the 2010 Census; my guess (and the guess of the U.S. Census demographers) is that the recent population increase skews in favor of Hispanic or Latino residents.

As of the 2010 Census, slightly more than 40% of the Harris County population identified as Hispanic or Latino, while about 33% identified as non-Hispanic white, and about 18.4% identified as non-Hispanic black. As of 2013, those percentages were estimated to work out as 41% Hispanic, 19% Black, and 32% non-Hispanic white.

A lot of people under 18 live in Harris County, so let’s say that only about 72 percent of the total population are voting age. That’s about 3,117,600 voting age people in Harris County as of 2013.

So … how successful have people been at registering Harris County residents to vote? Well, out of those 3.12 million voting-age people in Harris County, only 2.06 million were registered to vote in the November 2014 election. So let’s see, um, that’s a 66% registration rate.

Hmm. Let’s see. In 2002 (a relatively slow gubernatorial election year, with anemic turnout), there were 1.9 million registered voters in Harris County.

But … Wait. What? That’s almost as many registered voters as there are now in that county.

What was the voting-age population of Harris County in 2002? Well, the estimated 2002 population in the county was 3.54 million. Of those, we’ll guess that around 71% were 18 or older (based on the 2000 census data). So that’s an estimated voting age population of 2.51 million in 2002, of whom 1.9 million were registered. That’s a 75.6% registration rate.

I’ll admit that a 75.6% voter registration rate wasn’t spectacular in 2002. But it was definitely better than a 66% registration rate in 2014. If the Harris County Voter Registrar had managed to just tread water with regard to registration between 2002 and 2014, we’d expect that there would be about 2.36 million registered voters in Harris County as of the November 2014 election.

Which works out rather nicely as a shortfall of 300,000 people in the Harris County voting-age population who should have been registered voters (based on historic trends), but who weren’t registered.

Now … from 2002 to 2014, the racial demographics in Harris County basically flipped for white and Hispanic voters. In 2002, about 42% of the voters were non-Hispanic whites, and about 32% of the voters were Hispanic.

What the Texas Republicans realized in 2011 is that they would have to find a way to discourage both voter registration and voting in Harris County, so that while the percentage of non-Hispanic registered voters would remain roughly steady (rising a very modest 8% over the course of 12 years), the percentage of Hispanic registered voters would need to go down. And it would need to go down by a big margin.

The available tools were varied, and ranged from bad-faith purges and registration challenges to outright bullying and strong-arm tactics. But none of those tools would be as effective (in terms of the percentage of potential votes suppressed) as a simple increase in the opportunity costs of voter registration and voting.

This is an area deserving of additional study, but I would hazard that by focusing solely on the suppressive effect of voter i.d. laws as applied to registered voters, studies by the Brennan Center and others have dramatically underestimated the intended and successfully achieved levels of vote suppression generally. The underestimate is the result of failing to consider the suppressive effect of harsh election laws not just on the people who are already registered to vote, but also on the people who haven’t registered yet.

If the suite of laws enacted in 2011 (not just picture i.d., but also restrictions on volunteer deputy registrars, the mealy scrutiny and comparison of error-ridden databases, etc.) managed to drop voter registration in Harris County by around 13%, and the picture i.d. requirement managed to mess up voter eligibility for around 9% to 11% of the remaining 2.06 million voters, wouldn’t it be more accurate to say that (in Harris County, at least) that voter suppression succeeded in three ways, by:

(1) trimming the rolls of around 300,000 (mostly minority, poor, or elderly people) who would otherwise have registered to vote;

(2) discouraging another 180,000 to 220,000 registered voters from bothering to vote (the latter group also being disproportionately composed of minorities and the poor), and;

(3) so debasing and screwing up the election experience for everyone else that turnout was at an all-time low (around 678,000 votes cast in the big races, or around 33% of the registered voters).

It starts to look like the Republicans were running up the score in an unseemly way. I mean, if you start with 3.12 million possible voters (more or less), disenfranchise one sixth of those voters (say, around half a million), and then demoralize two thirds of the rest so badly that they don’t bother to come out and vote, well then … voila!

You get an instant Republican grand slam. A mandate for the ages, and an endorsement by our polity of the emerging πρωκτοκρατία (proctocratia, i.e., a form of government known as a proctocracy, administered by proctocrats).

Hey! You there! Candidates in Harris County who lost!

If you lost, and you lost by a margin of around 24% of the vote or less, strongly consider filing an election contest. Remember – such an action must be initiated not later than 30 days after the canvass of your election, by filing suit in district court. Your clock is ticking. Also remember (if you’re feeling a little sheepish about filing such a lawsuit) – that the November 2014 general election was an illegal election.

As for Mr.Cohn of the New York Times, here are some general tips for successfully navigating the confusing world we live in.

Sometimes bad actions are identified by absence, rather than presence. For example, if you are in a coal mine, and most of the canaries have suddenly gone silent, that isn’t proof that the remaining canaries are happy and content. That’s evidence that something is wrong.

“Well, everyone said that when the crew hit that natural gas pocket, it created a poisonous atmosphere. But I heard that at least 32% of the canaries were still singing. Therefore, the natural gas pocket could not have had any measurable effect on the health of the canaries. Q.E.D. Res ipse loquitor. Et cetera, et cetera. Say, does anyone else feel a little light-headed?”