By Cassidy Craford and Hannah Safford

Californians cite drought and water-supply challenges as some of the most important environmental issues facing the state today. A whopping 85% of California voters agree that water supply is a “big problem” or “somewhat of a problem” in their region. Population growth, dated infrastructure, and climate change are combining to strain water systems statewide.

The California legislature has long sought to mitigate water stress by passing water bonds that compensate for gaps in water-related funding. Such bonds were relatively modest throughout most of the 20^th century rarely exceeding $2 billion (in 2018 dollars) in any given year.[1]

The size of these bonds skyrocketed as California’s water problems became more pressing—and perhaps because raising money through a simple majority vote for a statewide water bond was politically easier than raising money through additional taxes or utility-rate increases. Despite the higher sticker price, the public remained on board. California voters have approved nine state bonds worth a total of $27.1 billion over the past 25 years.

That was the story until late last year. During the November 2018 election, California voters were asked to consider Proposition 3, an $8.9 billion water bond that would fund water-related infrastructure and environmental initiatives throughout the state.  Given California’s record of voter support for water bonds and expressed public concern for water safety and security, Prop 3 seemed likely to pass. But it failed by 0.65%—a slight yet impactful margin.

The obvious question is “Why did Prop 3 fail?” Multiple commentators have suggested answers. But exploring “Where did Prop 3 fail?” provides additional insights. The results are sometimes counter-intuitive…and deepen our understanding of how voters think about water in California.

Conventional wisdom predicts where water bonds should enjoy support

Multiple studies show partisanship affects how individuals vote on (ostensibly nonpartisan) ballot measures. Democrats tend to favor more government spending and greater government intervention than Republicans on nearly all issues—and environmental issues in particular. So we should reasonably expect water bonds to be more popular among Democrats, and more likely to succeed in Democratic-leaning areas.

This has certainly been the case in the past. In 2014, Proposition 1—a $7.1 billion bond to improve water quality, supply, and infrastructure—passed in all but one of California’s 20 most Democratic-leaning counties, but failed in half of California’s 20 most Republican-leaning counties. In 2018, Proposition 68—a $4 billion bond for parks, environment, and water projects —passed in all of California’s 20 most Democratic-leaning counties and failed in all of California’s most Republican-leaning counties.

Voter sentiment towards Prop 3 fell along similarly partisan lines leading up to the November 2018 election. Polling found that in July, 72% of likely Democratic voters were planning to vote “yes” on Prop 3, while only 43% of likely Republican voters said the same. But on Election Day, several counties bucked historical trends. Marin, Mendocino, Solano, and other strongly Democratic counties opposed the measure, while Republican-leaning Tulare and Kings counties supported it.

So if parties didn’t decide the fate of Prop 3, what did? A deep statistical analysis is beyond the scope of this piece, but we can look at four likely factors.

Factor 1: Funding priorities

The most obvious explanation is that not all water bonds are the same. Props 1, 68, and 3 had different funding priorities. For example, consider Central Valley voter interests in three measures. Props 1 and 3 earmarked millions of dollars for Central Valley water issues. Prop 1 allocated $34 million to the Tulare/Kern hydrologic region and $2.7 billion in funding for water storage projects (a traditional rural and recent Republican interest). Proponents claimed this funding would augment agricultural water supplies as well as help rural communities update failing wells and meet water quality standards.

Map showing the Friant-Kern Canal and other water infrastructure in Fresno, Kings, and Tulare counties.

Prop 3 proposed doing more. The Friant-Kern Canal sends water from Millerton Lake to San Joaquin Valley, serving approximately 18,000 farms and 160,000 families and businesses along the way. But, overreliance on groundwater has caused serious land subsidence which has damaged the canal and threatens future water availability. Prop 3 would have allocated $750 million to the Tulare/Kings/Fresno area, in part to repair and strengthen the Friant-Kern Canal. The sweeteners Props 1 and 3 offered the Central Valley were enough to get voters in Republican-leaning Tulare and Kings counties to go against party fiscal doctrine and support the bonds. But Prop 68—which did not contain such special measures—was predictably voted down in the region.

Factor 2: Timing

Countless researchers have demonstrated the importance of timing on policy success. This is illustrated by the fate of Prop 3. Not only was Prop 3 the most expensive measure on the November 2018 ballot, it also came quickly after Prop 68 in June 2018. Voters who historically supported water spending may have been turned off by another, larger financial ask arriving so soon (They want more money for water from me already?!). Moreover, the California drought was substantially over in 2018, compared to 2014. Absent an imminent water crisis, voters may have felt no need to support two water bonds in the same year.

Timing also may have affected voter confidence in the institutions that manage state water resources. One notable example comes from the region surrounding Lake Oroville, in light of the February 2017 spillway crisis. A survey of Oroville-area residents in October 2017 found 70% of respondents claimed that the Department of Water Resources had “lost all public trust with the communities downstream from the Oroville Dam.” Area voters who would have supported a $17.8 billion water bond in 2014 may have felt by 2018 that state agencies had proven incapable of handling such a large amount of funding. The Oroville data point could represent public sentiment more broadly.

Factor 3: Path to the ballot

Another problem with Prop 3 was its procedural history. There are two ways a state proposition can be put on the ballot in California. The first is by the state legislature—as was the case with Props 1 and 68. The second is by individual citizens through the “initiative process.” The initiative process allows any registered voter in California to propose a constitutional amendment or state statute, and for that proposition to go before voters on the state ballot as long as the proponent has a minimum number of supporting California voter signatures (typically 5% of the number of votes cast for governor).

One problem with the initiative process is that because California is so big and populous, the resources of at least one major institution or interest group are typically needed to fund the gathering of requisite signatures. The upshot is that “no initiative resulting purely from a volunteer drive has reached the ballot in three years.”

Opponents framed Prop 3 as an egregious example of “pay-to-play” politics that can shape the initiative process. Sierra Club California wrote that Prop 3 “[flew] in the face of good governance by being written behind the scenes.” The Sacramento Bee reported that California Assembly Speaker Anthony Rendon thought that Prop 3 was characterized by “a lack of oversight…and a surplus of special interest projects.” Such criticism may have swayed voters, especially those already turned off by the “insider baseball” side of politics.

Factor 4: The media

Media coverage is important in shaping election results. This is particularly true for the down-ballot policy proposals for which voters tend to have fewer preconceived notions. It is even more true in California, where voters are typically asked to consider more proposals than voters in other states.

Multiple studies have shown that even in the social media era, traditional newspaper endorsements can still significantly affect voter behavior. There is evidence that endorsements carry more weight when they (1) come from small, independent newspapers, (2) oppose public-benefit measures, and (3) concern economically relevant—rather than purely ideological—issues.

The theory suggests that media coverage helped determine who voted for and against Prop 3. The Bakersfield Californian and The Fresno Bee—two smaller newspapers serving Central Valley—both supported Prop 3. The emphasis that these papers placed on the specific regional benefits the bond stood to offer likely did much to win over even fiscally conservative voters in Tulare, Kings, and Fresno counties.

On the other hand, several papers serving Democratic areas strongly opposed Prop 3. The Sacramento Bee opined that Prop 3’s “list of beneficiaries” was “not enough to deserve voters’ support,” while the Marin Independent Journal underscored the problematic way that Prop 3 made it to the ballot. And indeed, neither Marin, Sacramento, nor neighboring Solano county ultimately supported Prop 3. Opposition from other papers—including the Los Angeles Times, the Monterey Herald, the Santa Barbara Independent, the San Francisco Chronicle, and the Mercury News—likely also did much to cut reduce support, even in counties that voted “yes” overall on Prop 3. Among the 20 most Democratic-leaning counties in California, county-level support for Prop 3 was, on average, 11.5% less than support for Prop 68 just a few months earlier.

The takeaway: Californians are willing spend on water…to a point.

California voters almost always say yes to bonds. Statewide bond measures pass around 90% of the time and California’s public has approved billions of dollars of water bonds specifically many times in the past.

But Prop 3’s failure proved that voters are willing to go only so far.  There is no simple way to predict when a voter will resist. Partisan lean is a predictor of the county-level vote on water bonds, but is by no means definitive. Funding priorities, timing, procedural history, and media attention are some of the factors that may also be at play.

Policymakers and bond proponents would do well to consider such factors. Voters have shown that they’re unwilling to just “throw money” at state water problems. To secure California’s water future, we must have policies to secure California’s vote.

Cassidy Craford is pursuing a B.S. in sustainable environmental design at UC Davis and is a student assistant at the UC Davis Policy Institute for Energy, Environment, and the Economy.

Hannah Safford is a Ph.D. student in environmental engineering at UC Davis and a researcher at the UC Davis Policy Institute for Energy, Environment, and the Economy.

[1] The major exception is a massive bond authorized in 1960 to finance construction of the State Water Project. This bond, though, was repaid with water-sale revenues, whereas recent water bonds are being repaid with general-fund revenues.