Library Referenda 2007: A Mixed Ballot Bag

According to the results of LJ’s latest annual referenda survey, covering measures held between December 1, 2006, and November 30, 2007, voters approved both operating and building referenda at very high rates. If your library reported a referendum this year, chances are you’re in high spirits. Operating referenda, with 69% of measures succeeding, enjoyed a good year, with more than $82 million approved out of approximately $103 million requested. Building referenda did a bit better, boasting a 74% passage rate and hitting a high not seen since the 2003 survey.

At the same time that passage rates trend upward, however, the actual number of referenda on the ballot continues to slide; building measures reported nationwide were down 22% since 2003, while operating referenda declined 65% during the same time period. What the data lets us surmise is that perhaps while libraries across the country recognize the general disposition of the public as unlikely to tax themselves voluntarily, we’re also getting a better handle on those ingredients that constitute a successful referenda campaign.

Even as libraries refine their ability to put together a package with a strong chance of succeeding on Election Day, events in 2007 further solidified the credibility of another piece of information that might guide referenda decisions. The average actual dollar amount per building measure last year was $13,186,892. However, with 46 separate building ballots reported going before voters, the 12 measures that were defeated asked for an average of $21,403,063. In fact, even removing from the equation the hefty $108 million, 30-year bond put forth by Sunnyvale, CA, referenda losers still averaged $13,530,614 per measure.

On the other hand, among the 34 libraries that acknowledged successful building referenda in 2007, the average proposal came in at only $10,287,068. It seems the size of the building referenda in dollars was the single most effective predictor of success this year, with a robust 80% of ballots under $10 million passing while only 63% of those exceeding that amount enjoyed success.

While the dollar amount of building referenda reported in 2006 said much regarding each referendum’s likelihood to pass and, likewise, conformed to the results of years past relative to this issue, building referenda bucked that tradition in 2007. Normally, building measures that go to the ballot between September and December have the greatest likelihood of passage; however, this year voters were feeling most generous from May to August. During this time period, 92% of the reported 13 building referenda on the ballot passed. Keep in mind, though, that even during the year’s stingiest period, from January to April, nine of 14 referenda got the thumbs up. With many wins garnering a two-thirds approval during any period of the year, the positive feelings the public demonstrated for building new libraries was underscored in 2007.

Neither was such strong voter support affected by the type of election. Proportionally, building measures were split evenly between general and special elections, and the results were virtually identical; 74% of the general election for capital funds were approved by voters, and a very impressive 18 of 23 special elections succeeded, for a hearty 78% approval rate. The unlucky few who did witness their building referenda fail were spread evenly across the nation and even among those failed ballots, support was reasonably high. Sunnyvale, CA, managing a very respectable 59% voter approval rate, still fell to the notoriously difficult supermajority requirement. Lake Orion Township Public Library, MI, and Lithgow Public Library, Augusta, ME, both came in only a few percentage points shy of the majority needed for approval. In the case of Lithgow PL, 48% of voters favored the proposed $6.9 million bond. Both the mayor and the local paper were vocal in their opposition, however, highlighting the importance of building support early on among local powerbrokers. Only those voters in Connecticut and Illinois felt less generous, with losing measures in both states topping out at 37% approval.

And, certainly, even as the compilation of statistics year after year provide for some solid clues as to what constitutes that “just right” mix of referenda variables, several successful campaigns highlight individual strategies. Viewing its referenda not as a goal in and of itself, Albert Wisner Public Library, NY, and Greensboro Public Library, NC, both recognize the ballot process as part of a much longer view. Rosemary Cooper, Wisner PL director, explains that the success of her ballot measure was based on “a strategy that reflected an excellent understanding of the community’s priorities and concerns.” Likewise, prior to the referenda proposed by Greensboro PL, the library solicited extensive input from the community in developing its Branch Improvement Plan (BIP). Thus, when the referendum was brought forward for consideration, the public felt it had a stake in the library’s direction. Using that deep community support as a base and suggesting that the Greensboro Friends use the BIP as a focal point in its own advocacy efforts led to very strong support at the ballot for a city the size of Greensboro.

Operating referenda voters came close in 2007 but did not quite match the zeal the public demonstrated for approving new library construction. While building referenda as a group averaged 59% positive support, the 29 operating referenda reported to LJ registered a still impressive 58% average “yes” vote.

Was there a common thread among those libraries whose operating ballots did not meet with voter approval? If so, it definitely wasn’t money. Unlike building referenda, the dollar amount requested in operating revenue apparently had little to do with the success of a given measure. The magic $10 million barrier didn’t stop the the two operating referendum that exceeded that amount in 2007; voters for the Toledo–Lucas County Public Library enthusiastically supported the $18.3 million ballot item, and San Francisco Public Library (SFPL) renewed its Library Preservation Fund with 74% approval.

The difficult times continued for Jackson County, OR, where voters emphatically rejected an operating levy for the second time in two years. The underlying problem for Jackson County is rooted in a larger county shortfall of $23 million owing to a failure on the part of Congress to renew a timber subsidy that provided Jackson County with significant revenue. The 42%–58% result was moot in the face of the daunting double-majority requirement for Oregon referenda that don’t appear on a general ballot. Apparently, however, one Oregon library failure contributed to another’s success. The Tillamook County Library braved the double-majority barrier, winning from voters an increase in the property tax mill rate on the same day as Jackson County’s defeat. What was behind the differing results? Tillamook director Sara Charlton during the campaign impressed upon voters that Jackson County Library had been forced to cease operations because of the previous year’s failed levy. Fearing a similar outcome, supporters in Tillamook County came out in strength at the polls.

Still another Oregon library that has experienced past difficulties with that state’s double-majority is Baker County Library District. In 2006, voters there demonstrated majority support for an operating levy but fell 60 votes shy of the required 50% registered voter turnout requirement. In 2007, in an effort to change its fate, Baker County placed the five-year local option tax on the general election and cleaned up with 59% of the ballot. Baker County’s success is not simply a matter of choosing the right date, though. Director Perry Stokes emphasized the library’s model of producing measurable results. By asking voters to approve short-term bonds and pointing out the service improvements those bonds foster, Baker County has built long-term credibility with its voters that translates into support at the polls.

Considering that failed operating ballots were, as a group, comparatively modest in their proposed increases to library revenue, it is not surprising that service population size correlated highly with voter approval. Of the 13 operating referenda put before the voters by libraries serving more than 25,000 people, only three failed. This 77% approval rate for mid-sized to large institutions is impressive. Conversely, of the 16 libraries serving fewer than 25,000 customers that placed a referenda item on the ballot in 2007, six were defeated for a 63% rate of success. This also correlates with the building referenda results; on that side, small libraries passed at an overall rate of 70% and larger libraries at 78%.

Library referenda planners often give considerable thought to the time of year that the vote will take place. Results from the operating referenda in 2007, quite different from our data in 2006, however, might serve only to heighten the confusion. Voters were clearly filled with the goodwill of the holidays in 2007 as ten out of 11 referenda presented to voters between September and December passed. Meanwhile, in 2006, this same time period was the least successful for operating referenda. Likewise, seven out of the nine failed operating measures in 2007 went before the voters between the months of April and June, a time frame that spelled success the previous year.

Two Northeastern libraries demonstrated the perils of having ballot funds rolled into a larger measure. The Edith Wheeler Memorial Library, CT, ballot passed, and the Whitinsville Social Library, MA, failed. In Whitinsville, town officials placed before the voters a $5 million levy to fund a wide variety of municipal departments. The measure failed with only 40% approval even though the library portion would only have been $210,000. Ann Dzindolet, the director at Whitinsville, remarks, “It was an all or nothing request, so the town voted for nothing.”

Robert Simon, Wheeler Memorial Library director, did get his money, but, because of the town’s annual budgeting process, the library received significantly less than needed to continue providing service in its brand new, greatly enlarged facility. Town administrators whittled the library budget down to level, equaling less than one percent of the overall town budget, while reducing this year’s request by more than five percent. Simon contends that the issue is “being lumped in with the town budget”; when the citizenry voted to fund the construction of the new library, it passed with 97% approval.

If the reported results from 2007 are any indication, being a piece of a larger measure was not helpful. What about “stand-alone” referenda on the ballot with other items? Overall, forcing voters to consider library referenda on the same slate with other ballot items produced mixed results. Over two-thirds of all passing building measures, 71%, were paired with other items. That figure was down to 67% of success among operating referenda. Without significant statistical data that favors or disfavors a single-item election, we polled some librarians whose measures did well on multi-item ballots, though their opinions also lack agreement.

Tobe Porter, director, Port Orford Public Library, OR, who needed high voter turnout to have a chance at meeting the double-majority requirements, felt that his library won “because so many people were against the other issue on the ballot” and turned out in high numbers to vote against it. Mary Beth Sharples, Midlothian Public Library, IL, director, believes high voter turnout in November 2006 owing to a contentious gubernatorial race doomed their measure. Trying again the following April, without any high-profile ballot items, Midlothian passed 51–49. Similarly, John Richmond, director at the Alpha Park Public Library, IL, felt that his library’s own levy proposal was defeated in the November general owing to significant turnout.

Alpha Park’s April special election was much more successful, garnering 64% voter approval. Some of that success, though, may come from taking a complicated financial proposition—shifting revenue generated by the millage dedicated to servicing construction debt into the town’s general fund—and simplifying the message. Instead of trying to explain this concept to voters, Alpha Park campaigned hard on the “voting YES means NO increase in the tax rate” slogan.

Public libraries that focused on getting out the vote had good results. Among the 46 libraries that brought a building referendum to ballot last year, most organized a campaign committee and the majority were victorious. Not all campaign committees are alike, though. In Peoria, IL, the library mounted a full-fledged publicity campaign, which included professional polling in advance of the election, 3000 yard signs, window signs for businesses, a web site, television commercials, political endorsements, and, best of all, the campaign committee marching in the local St. Patrick’s Day parade. The campaign was expensive, at over $100,000, but it looks like money well spent on a proposal worth $35 million. As a contrast to Peoria, both Lebanon Public Library, OR, and Otsego District Public Library, MI, went much more “on the cheap” but pulled it out nonetheless. When Denice Lee, Lebanon PL director, had only marginal success involving “key community leaders” to lead the charge, she and her library turned to a grass-roots effort of “worker bees” and spent only $15,000 in order to win millions. At Otsego District PL, Director Ryan Weber did even more with less. The campaign in support of a $3.6 million bond cost the committee only $500. With the library squeaking by on a 51–49 margin, that effort may have made all the difference.

However, well-organized campaign committees can work for the benefit of those opposed to a given ballot item, too. Two libraries that issued bond proposals for new construction learned this the hard way. Loutit District Library, Grand Haven, MI, was victorious in 2007, picking up 64% of the vote; however, in 2004, a similar, slightly larger bond proposal was defeated by the same margin on the strength of “a very well-organized opposition.” Carol Stream Public Library, IL, met with “a very strong and nasty antireferendum campaign” this year and garnered only 30% support as a result.

Public libraries that went before the voters with proposals to increase operating revenue were much less likely to mount the kind of publicity effort common on the building side. And, frankly, results for operating referenda in 2007 were no less the worse for it. The difference separating the percentage of “yes” votes among those operating referenda proposals that did feature a campaign committee and those that didn’t was, astoundingly, less than one-thousandth of a percentage point. The Lapeer District Library, MI, illustrates that sometimes “facts on the ground” are just as motivating as a good campaign. Yes, Lapeer did run, by all accounts, a quality campaign for levy passage last year, reaching out to virtually every segment of the community; however, the campaign itself might never have gotten off the ground had not the individual who eventually organized it come to one of Lapeer’s branches only to find it closed owing to reduced hours resulting from the town’s previous, expired millage.

Overall, our 2007 survey was marked by strong support for public libraries on Election Day. Several proposals across the nation, however, did not succeed, and those libraries’ futures are unclear. A respondent from Manhattan Public Library District, IL, the only library this year to be voted down both in its attempt to raise construction funds and for an increase in operating revenue, states plainly that there is “no choice but to try again”: the town faces a projected boom expected to quadruple the service population in only 15 years. Sugar Grove Public Library District, IL, like Manhattan, feels as though its hand is forced to bring another vote to ballot. A new facility there is already under construction; an increase in operating revenue is a must. Jackson County, OR, buoyed by a one-year extension of timber funds, has outsourced operations of the library system to Library Systems & Services, LLC (LSSI). The additional revenue should allow the library to operate for three years on a sharply reduced schedule while library stakeholders search for a long-term solution.

It is difficult to predict, though, if these libraries and others like them will enjoy a better fate at the ballot in 2008. Nationwide, as voters witness city and county budgets tighten because of revenue shortfalls and worries spread of a downturn in the economy, it is unlikely that the public will be willing to add to its tax burden. Here’s to the hope that the public will maintain their high opinion of our libraries as difficult choices are made in the year to come.