Sounds like a common challenge.
I’ve also experienced this. So, I’ve been focusing more on the making the case for ‘lowering the bar’ in terms of moving public agencies toward transparency. Public officials do have legitimate concerns regarding data accuracy and privacy issues, the value of transparent data for citizens, as well as political reverberations and resource challenges. Nevertheless, I believe these are some of the obstacles that need to be navigated through on the way toward open data.
To me, the “lowest bar” is the most basic use case: publish data on the web in a raw format (like .csv) on a regular and timely basis. This is technically very feasible through scripts and a scheduled task/cronjob, with FTP to a public site. Once this feasibility is established, other concerns may arise: privacy, data integrity (accuracy), questioning the value of the data to the citizens*, and primarily the uncertainty that comes in the public arena for starting something new**
Regarding privacy, it is always helpful to have legal counsel review data before it goes live, and better yet, develop a policy for releasing data (a policy may already be in place for FOIA requests). Basically, remove non-public, personally identifiable information. Also, consider ‘sensitive’ datasets in terms of safety and security (politically sensitive doesn’t count).
Regarding data integrity, if the information is good enough to use inside the organization, it is good enough to publish. I think SF has an interesting case study on this where a member of the public identified some discrepancies with Property Tax data. Yes, a bit embarrassing, but this is exactly the value of open data: improvement.
* Sometimes this is communicated as: “citizens would not know how to use the data, we should make it into a report instead” – which I respond that every manipulation of the data is a biased view. And the data should come with the same spec in-house developers have (although this is sometimes a proprietary system/contract issue). Plus, making a new report raises the bar, and serves as an obstacle to truly open data. Additionally, it is helpful to reference the ecosystem of opengov developers who are willing and able to help make sense of this raw data, which also helps reduce the burden on government themselves.
** Case studies are useful for conveying the fact that open data is not brand new. There are several agencies that are doing this successfully and seeing beneficial results for doing so (SF, DC, Seattle, Portland, Manor TX, Oklahoma, Vancouver, and many more places in Europe).
Good luck in Petaluma! I’m facing the same challenges in Solano County, with varying degrees of success.