Improving the CIP Fiscal Note Process

During the 81st Regular Session, the volume of requests from the LBB was much, much greater than in any previous session. The County Information Project (CIP) worked on 220 separate requests from the Legislative Budget Board – and we anticipate that the 82nd Regular Session will be even more hectic! Technological advances continue to allow more and more bills to be filed each session. In addition, by federal law, the Census Bureau must release redistricting data from the 2010 Census by April 1, 2011. These two factors indicate that we can anticipate an increased work load during the next session.

During the past session, it wasn’t just CIP staff working on those replies. A recent article in the TAC staff newsletter thanked 23 employees by name for their assistance with those replies.

Given the amount of effort involved, it is imperative that we review our work processes and capture our own “best practices” in order to prepare for 2011. When it comes to our fiscal note process we already recognize the need to increase the participation of county officials.

There are many bills filed for which we just do not have adequate data or knowledge to respond directly to the LBB. In those cases, we have to send out our own information requests to counties in order to determine a response for the LBB. It is almost always the case that we get only a small number of responses; typically only from those county officials who are highly motivated by that particular bill. More than one official told us point-blank that they will not respond unless a bill affects their county significantly.

This practice presents us with a bit of a problem. It is not uncommon to contact a dozen counties about a bill and only receive one or two responses. Does that mean that the remaining counties foresee no significant impact or are the officials just too busy to respond? If one county says the bill will cost nothing and a second responding county says the same bill will cost them a million dollars, how do we determine the statewide impact on counties?

If time allows, we can send out a second information request. Given the time constraints, that is generally not an option when working on fiscal notes. As a result, although there is no hard and fast rule, we’ll try to extrapolate to the 254 counties if we have more than a handful of responses. It really comes down to a judgment call. We have to ask ourselves two questions. First, do we have a sufficient number of quality responses that are in agreement, more or less? Second, how do the non-responses affect our response?

If we are truly uncomfortable with the number of responses we merely report the data to the LBB. That doesn’t really solve anything; it just moves the burden from our shoulders to theirs. Of course, typically the LBB analysts then merely add our information to the fiscal note which will then state the impact on those few counties rather than to counties in general.

But what should we do with the non-responses? Our practice this past session was to ignore them; there was really no other option at the time; but now the session is over and there is hopefully some time before the next one starts. There are actually a number of reasons why this issue is of some concern.

Not responding when a bill will have little or no impact skews the statewide extrapolation. For example, we receive four responses after attempting to contact twelve counties. All four responses come from counties with a population over 150,000 each of which states the bill will have significant fiscal impact. What will the impact be to small counties?

That’s a hard one, let’s try an easier question. What will be the statewide impact on counties? One-third of the contacted counties say the bill will have a significant impact. But, two-thirds of the counties may not have responded simply because they didn’t see any significant impact from the bill.

By ignoring the non-responding counties, are we throwing away data? Clearly, by failing to obtain responses from all of the contacted counties, we have introduced a statistical bias into whatever answer is provided to the LBB.

We exacerbate that bias when we rely on the same counties or same county officials over and over for assistance. For the most accurate fiscal note, we should choose the counties at random. But doing so provides a very poor response rate. As a result we tend to overburden some county officials in order to have at least some usable data.

If you think those fiscal note questions are tough, here’s one that will really make you think. If we get widely varying responses from counties, who do we believe? What if one county says this bill will be very expensive while another says there is no way this bill could possibly have a significant impact? What if we get two responses from the same county and the two county officials have widely diverging estimates of the bill’s impact? Should we attempt to verify responses? Which ones?

The question is largely academic; during a session there simply is no time to verify responses, we have to accept the data at face value. Even if there was time, there is no reasonable method that we can see to verify responses. The best answer we have found so far is to obtain data from as many counties as possible so that individual responses, no matter how extreme, have less influence on the overall fiscal note.

If you think those are all hypothetical questions about our fiscal note process, think again. They all came up at one time or another during the session.

The LBB seemed satisfied with our work product this session – they sent a very nice thank you note. We think we should be able to do better – we can do better – but only with your continued support.

Please contact me at TimB@county.org if you have any questions or comments about this article. I’d appreciate hearing from you, but don’t expect a quick response; I will be away for two weeks in June on active duty fulfilling my annual commitment to the U.S. Navy Reserve.