How Research Set the Stage for the Overwhelming Defeat of a Measure that Previously Got 66% Yes

Over the last several weeks, we’ve reviewed the six phases of a HexaCom Group campaign, our approach to each, and how they work together to win difficult campaigns.

What we want to do next is show some real case studies of our approach as applied in campaigns we’ve run. In this case, the No on Measure U campaign in Adelanto, California on the primary ballot in 2024.

And we’ll put the TL/DR here upfront, although we hope the full article is worth your time:

1. Research is more than opinion research.

2. You need policy research, previous voter histories, coalition mapping, etc, before you do your survey.

3. If you are running a campaign against a ballot measure – READ THE BALLOT MEASURE! You will find avenues for attack.

4. Get professional policy and opposition research if you can afford it on your campaign, but do it yourself if you have to.

5. If there are relevant previous elections, analyze those for any insights into voting patterns.

6. Analyze the coalition playing field with an eye toward who might weigh in the campaign on your side and the other. Yes, this takes research. Don’t waste time testing in your survey the impact of a community leader who will never weigh in on your issue. Nor testing a statement they never would make.

7. Pick a good research partner for the campaign, and work collaboratively with them to ensure they understand your research goals you have for the project, and that they have all the information they need to draft a good survey questionnaire.

Getting back to the campaign proper:

Measure U was the City’s second attempt to pass a broad “vacancy” tax, which would have increased property taxes (via a schedule of parcel taxes) on what they defined as vacant, unused, and/or under-utilized land. The proposal detailed a specific per acre parcel that would be imposed under certain circumstances deemed “vacant,” and specified how the $6.2 million raised from the taxes would be spent, including public safety, infrastructure, mitigating blight, and park & recreation. Under California law, this is considered a “special tax” and requires a two-thirds vote to be adopted.

The prior effort to pass a vacancy tax in Adelanto in 2020, Measure R, fell just short of the two-thirds majority, receiving 65.7%. However, no significant campaign had been run against the 2020 effort, so we knew there was support for the concept of a vacancy tax in the City, but it was not clear how deep that support was.

The city chose to try again in March, 2024, under Measure U. We were hired to run the campaign late in the process, just three months before election day.

And since it all starts with research, we dove right in, and we had to move fast, because coming out of the Holidays, we knew we’d have an eight-week sprint to Election Day.

To many folks, research means just opinion research, but as we explained before, research is multi-faceted, including policy research, past election results, the likely coalition playing field, and more. All of that research provides us with the information we need to ask the right questions in opinion research.

Here’s one important take away, especially when you are on the NO side of a ballot measure – READ THE MEASURE! It seems simple enough, but it is so important. You must read the measure – line by line – with a keen eye toward vulnerabilities, unintended consequences, and other avenues under which the measure can be attacked.

We can’t tell you how many times we’ve known we were going to win a campaign as soon as we read the opposition measure.

In the case of Measure U, we did just that, as well as the materials and the statements the City was using to justify the proposal.

As you can see in the attached, we went through the Measure itself (which is after all what the voters are being asked to adopt), the City’s official argument in favor of the Measure (as filed by the Mayor,) and the City Attorney’s non-partisan analysis of the measure.

We were also very fortunate to have a thorough legal analysis of the Measure done by the National Association of REALTORS that helped us identify potential flaws and problems in the text of the ordinance itself.

Potential avenues of attack immediately jumped out to us.

The City said they needed this $6.2 million, or vital public services would be cut. They had said the same thing back in 2020 when they tried this before.

So, we downloaded the City’s last four budgets, since the failure of Measure R in 2020, to see if the City really “needed” the money. And guess what we found? The City’s budget had gone up since the last time they said they “needed” the money. We also looked, budget category by category, and found one of the areas of fastest growth was how much the City Council spent on themselves. Virtually every category that Measure R was intended to fund had seen budget increases since Measure R failed.

That fact definitely got filed away for later use!

We also noted that while the City sold the measure as having strict accountability on how the money would be spent, they in fact gave themselves wide latitude in how they could actually spend the money after the first year...in other words, the city’s promise on how to spend the money was only good for one year, and for years 2-20 of the twenty year tax, they had a lot of ability to move the money around.

Another fact found directly from the review of the text of the measure to best tested.

Another key argument the City was making is that only “out of town” land speculators would pay the tax, not locals.

So, we pulled the parcel list of so-called “vacant” properties, and while a majority of parcel owners were not residents of the City, hundreds of parcel owners did live in the City. What’s more, some of the parcels deemed vacant were vacant for important reasons, such as sensitive habitat, yth Measure R would still tax them extra.

Mark that one down too!

Additionally, the City argued that the tax would only apply to vacant lots, not lots with buildings, but when you read the actual text of the Measure, it was clear that even if a developer put a building on a vacant lot, if no one occupied that building under the strict guidelines in the measure of what constitutes “occupied,” it would still pay the tax. A developer could build a retail center on a lot they owned, but if no tenants came, Measure U still counted that property as vacant.

Check.

Finally, we noted with the help of policy and legal analysis several circumstances in which it might be common for someone to not occupy their property, such as military service, illness, a religious mission, and caring for an elderly parent. All of those circumstances could lead to someone having the building they occupy – i.e. their home – to be deemed “vacant” by the City and subject to extra parcel taxes.

Definitely on the list of things to test!

Finally, in reading the measure, we also saw a couple issues with the attempt to show voters that there would be oversight and accountability on how the new taxes would be spent. First, that the so-called “Independent Oversight Commission” had no authority, were put in place by the City Council they were supposed to oversee, and that they could be fired by the City Council. Second, the measure had no “Maintenance of Effort” provision, which means nothing in measure prevented the Council from reducing funding currently allocated for programs, backfilling with Measure U funds, and then spending money however they wanted.

At this point, we felt like we had a good inventory of potential arguments to test for an against the measure, but we had a bit more research to do.

We took a good, long look at the results from the Measure R campaign, looking at the areas in which the Measure had done relatively better or worse, and drawing some conclusions from which demographic groups were likely more or less supportive.

And lastly, we wanted to know which organizations would likely weigh in on the Measure, either for or against, and what they might say, so we could test realistic arguments in our survey. Notably, we needed to know whether the Sheriff's Employees' Benefit Association (SEBA – which is effectively the San Bernardino County Deputy Sheriffs Association) would weigh in favor, since the Sheriffs Department provides police services to the City of Adelanto by contract. We also reached out to local community leaders, and organizations, such as taxpayers’ association and the local Chamber of Commerce. Fortunately, it appeared SEBA was going to sit this one out, and voices of the small business community and taxpayers’ associations would be on our side if we wanted them.

Because we were under time and budget constraints, and we wanted to get all research done before the Holidays so we could start the campaign right after the New Year, we skipped qualitative research and went straight to getting our survey done before it got too close to Christmas.

On this campaign, we worked with Trevor Smith of Pulse Decision Science (still known at that time by their previous name, WPA Intelligence) on this project.

As a side note, the process by which we pick the right opinion research partner for a project is varied. Yes, we have firms we really enjoy working with and know that they do great work. Sometimes it is experience in a location or an issue area. Sometimes the client may have a preference. Last year, we worked with many different research partners. For this project, we faced a time crunch, and the challenge of getting a valid sample size in a small town like Adelanto, and we’d successfully faced those challengers with Trevor before, and it was the right call.

Drafting the questionnaire for a survey is a collaborative process between us and the pollster. We will often do a research memo to the pollster on our goals, what we know, and what we want to test. They will then draft the questionnaire, and we’ll go back and forth until we’re happy with the document, and then field it.

In this campaign, after initially pressing to get the survey in the field and done before the holidays, we ended up waiting until early January to field. We had too much concern, given the small voter universe and the need to get good data in a tough environment for polling not to risk trying to field in the weeks leading up to Christmas.

The survey accomplished everything we hoped it would, showing as we suspected the near 66% support that Measure R had in 2020 was broad but not strong or deep, and that the measure was very vulnerable.

We’ll get deeper into the results of the survey and the strategic insights we derived from it in our next entry.