Analysis of the 2004-05 Budget Bill
Legislative Analyst's Office
The 2003-04 Budget Act abolished the Technology, Trade, and Commerce Agency (TTCA) and state funding for many of its programs designed to encourage economic development. The state's adverse fiscal condition precipitated the elimination of the agency. Even during better fiscal times, however, we had long questioned the value and impact of many agency programs. Indeed, even at its peak level of funding, TTCA's programs assisted a very small portion of targeted industries or businesses. In other words, almost all economic activity in California occurred without the agency's direct intervention through various programs providing services or subsidies to businesses.
The Governor's budget does not propose any restored funding for these programs. In the future, however, the Governor's budget or other legislation may propose restoring some of these programs or creating new ones to promote economic development. In this piece, we highlight some issues for the Legislature to consider when reviewing any future economic development proposals.
The state plays a critical economic development role in establishing the social and physical infrastructure for business activity and economic growth. Two of the most prominent and important examples are (1) the K-12 and higher education systems to provide an educated workforce and (2) the transportation system to promote the efficient movement and provision of goods and services.
State involvement should be unnecessary in most individual business transactions. Exceptions can arise in the case of "market failures," in which transactions among businesses produce less than efficient, or optimal, outcomes. For instance, a number of federal and state programs target small businesses because it is believed that they are often not able to compete or procure services on a level footing with larger companies having more resources at their disposal. Other situations may arise where state involvement may be justified due to concerns regarding the distributional results of market conditions which are not viewed as appropriate. For example, certain economic development programs may be specifically directed towards areas with high rates of unemployment and poverty.
Extensive academic studies have evaluated the effectiveness of local, state, and federal government economic development programs. These studies separate out the effects of many variables on economic development or business activity. On occasion, a comprehensive study will conclude that a program has a measurable, positive impact on business activity or economic development. They often find, however, that (1) particular government programs have a small impact, if any, on the measures studied—employment or sales growth, for example—or (2) they have a poor cost/benefit ratio.
Often lacking the data, time, and expertise to perform such complex studies, the state must typically rely on less comprehensive evaluations. However, these analyses can still lend significant insights into program outcomes, lead to reasonable conclusions about program effectiveness, and serve as the basis for good policy decisions.
Determining Net Impact Is Critical. In the past, our reviews of TTCA programs found that the data offered by the agency often did not accurately address program effectiveness. The fundamental question of how recipient businesses would have performed in the program's absence was often unanswered. For example, if new sales or jobs would have occurred without the program, then those results cannot truly be claimed as an outcome or benefit of the program. Data that accurately demonstrate pro gram impact would account for this underlying growth and only focus on additional business activity among assisted businesses.
As an example, the cost-benefit analyses the agency completed for the state's foreign trade offices counted the entire value of transactions in which the state was involved as a program benefit attributable to the offices' presence. This conclusion assumes that absent the state's intervention, these exports and foreign investments would not have occurred otherwise. That is, the agency claimed responsibility for the full value of all these exports—an assumption that overstated the value of the offices' involvement with these transactions and, more generally, direct state intervention in foreign trade.
As an alternative, to determine a better estimate of the effectiveness of the foreign trade offices, the agency could have asked assisted businesses about the impact directly attributable to program assistance. (This is done in an annual survey for the Manufacturing Technology Program, which provides business consulting services to small manufacturers.) In addition, TTCA could have considered overall trade-related activity among similar businesses and/or with the respective foreign countries. This would have presented a more accurate picture—one that highlighted any observed differences between underlying business activity and that of the businesses assisted by the trade offices.
We recommend that the Legislature require a well-researched problem, a demonstrable net benefit, and built-in evaluations when considering any future proposals for state economic development programs. This would allow the Legislature to be more certain of the effectiveness of proposed programs.
As discussed above, the most comprehensive academic studies of economic development programs typically show a lack of quantifiable results. As a result, we recommend that the Legislature set a "high bar" when considering any future proposals for state economic development programs. If a proposed program appears to "pan out" with such an approach, then the Legislature could be more certain that the program would have a net impact on business activity. Specifically, as shown in Figure 1, we recommend that the Legislature require the following components in proposals for economic development or trade-related programs.
An identified, researched problem in the private market, and an
assessment of how state action would address the problem.
Benefit-cost ratio greater than one-to-one, without
Built-in data collection and program evaluation focused on
An Identified, Researched Problem. Proposals—whether programs that are new or reconstituted from the former TTCA—should include a thorough explanation of an identified market failure leading to the proposed state intervention in the marketplace. In addition, it needs to be shown how state intervention would actually help address or mitigate the identified market failure. The proposal should include data and research supporting the conclusions. Such analysis could come from the state's Economic Strategy Panel (responsibility for which has been transferred to the Labor and Workforce Development Agency as a result of the TTCA elimination), relevant data from previous TTCA programs, or data on similar programs sponsored by other states or the federal government.
More Than a One-to-One Return. Programs focused on increasing business activity should, in theory, affect state tax revenues. In fact, TTCA program evaluations often presented data claiming additional state tax revenue in excess of program costs to demonstrate program effectiveness. Given the prevalence of extensive academic studies showing minimal or inconclusive impact from such programs, however, a ratio of new revenues generated by the program compared to program costs that is appreciably greater than one-to-one would provide a "cushion" for the Legislature to be more certain that the proposal would have a net financial benefit. (While there are certainly benefits other than tax revenues from economic development programs, most programs have tended to rationalize their worth using net revenue impacts.)
To enhance confidence in this cushion, no job or other multipliers should be used in evaluating program effectiveness. A recent report from the federal government indicates that job multipliers developed from macroeconomic models of the economy (at a scale of trillions of dollars) are not sensitive enough at the microeconomic level to meaningfully apply to programs aimed at individual businesses (at a scale of millions of dollars or less). In short, the multipliers are not accurate on a small scale. If such a standard of not using multipliers were adopted, the Legislature could be more confident that a program estimated to have a direct, positive impact on targeted business activity and/or state tax revenue would truly be able to achieve this result.
Built-In Evaluations. Proposals should document that the programs will account for baseline business activity in some manner. As a result, any approved programs would automatically be generating data and analyses for future use in reviewing program effectiveness.