The political action committee (PAC) that's pushing the referendum that asks Escambia County citizens to surrender their right to vote for their superintendent has been telling voters that 99.5% of the nation's school districts appoint their superintendents. They use that data point to tell us that since virtually all other school districts in the country appoint their superintendents, Escambia County's method of selecting its superintendent by voter ballot is antiquated and "archaic."
The impression that we're left with is that nearly all of the school districts across the nation are foregoing the "elected" format in favor of what the PAC says is the more enlightened "appointed" format. We citizens then assume from what they're saying that school districts nationwide are fairly comparable in their structure, size, composition, and support apparatus, and that the chief difference between the Escambia County School District and 99.5% of the others is the method by which we select our superintendent. They lead us to the conclusion that it is that factor - changing the method by which we select our superintendent - that is the prime catalyst for education growth and success.
It turns out, however, that's not the case. They're comparing apples to oranges.
States that include California, Arizona, and Montana, for instance, have a two-tiered district structure in which citizens ELECT the overwhelming majority of their county superintendents. The county superintendents oversee appointed "local" district superintendents who preside over smaller geographic areas of the county. Thus, the roughly 1,700 local district superintendents in those three states are appointed by their local school boards or county boards, but most of the superintendents who oversee education at the county level in those states are elected, as is the case here in Escambia County.
Looking at California, 53 of the state's 58 county superintendents are elected by the citizens to oversee the state's "local" school districts, which number around 1,000 school districts for the state's 58 counties. 
Although there are 53 elected county superintendents in California, if a civil grand jury in Santa Clara County has its way, that number will rise to 54. Just 4 months ago, the grand jury recommended that the county change from an appointed county superintendent to an elected one in part because of the high turnover of appointed superintendents there: 5 superintendents in the past 11 years. The grand jury also found that "an inherent tension exists between appointed superintendents and the appointing BOE" (board of education). The finding also stated that the Santa Clara County school board had often exceeded its authority by encroaching into the day-to-day management of the school district which is the superintendent's domain. The report observed that the challenge for the superintendent, who has independent constitutional and statutory authority as he or she does in Florida, is the need to appease the board which has the authority to terminate the superintendent's "at will" employment. 
Like California, Arizona and Montana also have county superintendents who are elected by the people to oversee local school districts. The people of Arizona elect 15 county superintendents and Montana voters elect 56.  
Then, there's New Jersey which also has county superintendents, 21 of them. However, they are all appointed by the governor with the advice and consent of the state senate. New Jersey's more than 500 local superintendents report to them. 
There are variations of that oversight structure in other states like Illinois which has 56 regional superintendents who oversee their more than 800 local school districts and superintendents. Illinois' regional superintendents are elected to their positions by the citizens. 
Unlike those states and others that are similar to them, Florida school districts - including the Escambia County School District - do not have subordinate "local" school districts within their large county systems. The absence of the additional structure above and below Florida's large county-wide school districts also explains why Escambia County's school district, with its 40,000 students, is larger than 99% of school districts nationwide, one of several reasons that objective comparisons of hierarchical structures are difficult to make. Here in Florida, more than 61% of the school districts elect their school district superintendent.
Additionally, most states have regional cooperative educational service centers or alliances that provide a range of cooperative educational services to groups of local school districts so the local school districts can focus more on classroom education, but Florida school districts do not. The fact that most U. S. school districts are able to leverage that kind of resource is significant to their operations and organizational structure.
The table below illustrates the effect of that smaller local structure on the scope of school district work at that level. It reveals that the five states mentioned earlier - California, Arizona, Montana, Illinois, and New Jersey - all benefit from their structure in that they have a higher average number of school districts per county and a lower average number of students served by their school districts than Florida has. Also included in the table is data for Wisconsin which has been referenced by supporters of the referendum for comparison. The contrast between these states and Florida in this regard is striking.
So, when the PAC that's pushing this referendum says that 99.5% of school districts have appointed superintendents, they're trying to compare apples to oranges. The structure, size, composition, and support services apparatus here in Florida are vastly different than they are in every other state. We know that in the real world of school district operations, those and many other factors matter.