NRC Panel Publishes Report on Productivity Measurement in Higher Education
A few weeks ago, the National Research Council’s Panel on Measuring Higher Education Productivity published its 192-page report on Improving Measurement of Productivity in Higher Education, marking the culmination of a three-year, $900,000 effort funded by the Lumina Foundation and involving 15 higher education policy experts nationwide.
In explaining the need for a new productivity measure, the Panel made several key observations:
It’s all about incentives: Institutional behavior is dynamic and directly related to the incentives embedded within measurement systems. As such, policymakers must ensure that the incentives in the measurement system genuinely support the behaviors that society wants from higher education institutions and are structured so that measured performance is the result of authentic success rather than manipulative behaviors.
Costs and productivity are two different issues: Focusing on reducing the cost of credit hours or credentials invites the obvious solutions: substitute cheap teachers for expensive ones, increase class sizes, and eliminate departments that serve small numbers of students unless they somehow offset their costs. In contrast, focusing on productivity assesses whether changes in strategy are producing more quality-adjusted output (credit hours or credentials) per quality-adjusted unit of input (faculty, equipment, laboratory space, etc.).
Using accountability measures without context is akin to reading a legend without looking at the map: Different types of institutions have different objectives, so the productivity of a research university cannot be compared to that of a liberal arts or community college, not least because they serve very different student populations who have different abilities, goals, and aspirations. The panel notes that, among the most important contextual variables that must be controlled for when comparing productivity measures are institutional selectivity, program mix, size, and student demographics.
The Panel also contributed a thorough documentation of the difficulties involved in defining productivity in higher education. From time to time, it is helpful to remind ourselves that, while it may be “possible to count and assign value to goods such as cars and carrots because they are tangible and sold in markets, it is harder to tabulate abstractions like knowledge and health because they are neither tangible nor sold in markets”. The diversity of outputs produced by the institutions, the myriad inputs used in its activities, quality change over time and quality variation across institutions and systems all contribute to the complexity of the task.
Despite these difficulties, the Panel concluded that the higher education policy arena would be better served if it used a measure of productivity whose limitations were clearly documented than if it used no measure of productivity at all. It proposed a basic productivity metric measuring the instructional activities of a college or university: a simple ratio of outputs over inputs for a given period. Its preferred measure of output was the sum of credit hours produced, adjusted to reflect the added value that credit hours gain when they form a completed degree. Its measure of input was a combination of labor (faculty, staff) and non-labor (buildings and grounds, materials, and supplies) factors of production used for instruction, adjusted to allow for comparability. The Panel was careful to link all components of its formula to readily available data published in the Integrated Postsecondary Education Data System (IPEDS) so that its suggested measure may easily be calculated and used. It also specified how improvements to the IPEDS data structure might help produce more complete productivity measures.
The key limitation in the Panel’s proposal – fully acknowledged in the report – is that it does not account for the quality of inputs or outputs. As the Panel notes, when attention is overwhelmingly focused on quantitative metrics, there is a high risk that a numeric goal will be pursued at the expense of quality. There is also a risk that quantitative metrics will be compared across institutions without paying heed to differences in the quality of input or output. The report summarizes some of the work that has been done to help track quality, but concludes that the state of research is not advanced enough to allow any quality weighting factors to be included in its productivity formula.
While readers may lament the Panel’s relegation of measures of quality to further research, especially given the time and resources invested in its effort, the report remains a very useful tool in understanding the issues involved in assessing productivity in higher education and provides valuable food for thought for policymakers and administrators alike.