Productivity vs. Performance Measures

I went to a high-powered panel presentation last week sponsored by the Center for American Progress on “Government’s Productivity Imperative” and it left me a bit nervous.  It was inspired by a McKinsey article on productivity and it urges the Obama Administration to adopt productivity improvement approaches that could save $45 billion to $134 billion a year, noting “. . . an imperative for dramatic productivity improvement in government that will delivery better results for citizens.”  A companion McKinsey report noted that the federal government last measured its own productivity in 1994 and hypothesized that government productivity lags private sector productivity.

The McKinsey authors believe that reinstituting productivity measures would create public pressures to improve productivity and save money:   “Measure productivity again, providing an ongoing, transparent indicator of progress. Set an ambitious national government productivity target to match private sector rates of growth.”  The Washington Post breathlessly called this a “Productivity Revolution.”

I’m not as confident that this would be the case.  Doing so would take years and millions of dollars to reinstitute the measures.  Government could do better by emphasizing improved performance and results, instead.  It already has that information.

In fact, the McKinsey article observes:  “Making the performance of governments more transparent by publishing the results of customer satisfaction surveys, benchmarking surveys, and service-quality metrics also help citizens to take an active role in demanding change and applying appropriate pressure for performance improvement.”

Why return to the days of the Carter Productivity Improvement Programs and arbitrary targets of 10 percent productivity improvement in each agency, accompanied by savings demands that may or may not be possible?

What is productivity?  According to the McKinsey researchers, “Productivity is the amount and quality of the goods and services that can be generated with a given set of inputs.” 

Why doesn’t government measure its own productivity?  The Bureau of Labor Statistics used to measure federal productivity but stopped collecting this information in 1996.  At that point in time, productivity lagged private sector and McKinsey’s report concluded “our hypothesis is that, since then, this gap may well have grown.”   They then base their projected savings on this hypothesis.

I was involved in terminating government productivity measures when I was working for the Reinventing Government initiative.  The methodology was specious – largely self-reported data on administrative processes and not government mission-related work.   There was also a 2-year lag in the data reported, so it was largely useless to make decisions.  More importantly, government could not measure its inputs.  And even 19 years after the Chief Financial Officers Act was passed, government still has no activity-based costing so it can measure inputs.  This is a fatal flaw in measuring the inputs for a productivity measure.

There are also problems with measuring the amount and quality of government services produced.  Some economists conclude that in a service economy, especially the government, “the concept of productivity has become meaningless.”  They say productivity measures capture only one dimension of performance – that performance can be assessed by different stakeholders with different sets of values.  In some cases, equity and fairness is as (or more) important than cost efficiency.  In addition, there is controversy among economists as to how to best measure productivity of services that are unpriced, like government services.  How do you measure the productivity of firefighters?  By the number of fires they put out?  What if they are successful at preventing fires? Does that mean their productivity drops? . . . What about the productivity of national defense? . . . how do you account for a service that produces a “collective good?”

Fixing these technical problems with productivity measures would have cost millions and take years to implement.   It seemed better to kill them because occasionally the data were used to set arbitrary savings targets (see the Reagan productivity initiative) that led to a waste of time and energy.  One veteran of the process called it “a fool’s errand.”  And there was the fear that targets would be set politically, savings claimed, then it would be up to agencies to find ways to cut their budgets to meet these productivity targets.  No wonder productivity had such a bad name in government!

Other countries have developed better measures of government productivity, but the question in the U.S. is, is it worth the time and cost if there are other approaches to get at the same end?  This is hinted at in the McKinsey report.

Still, there is value in performance transparency.  The McKinsey report reaches beyond its initial emphasis on productivity.  It also notes that it is important to create more transparency in agency and program performance through greater comparability, accessibility and independence in the scrutiny of government data.  This can be done – and be agency- and mission-specific – using performance measures to assess progress and results.  This doesn’t require productivity measures.

In fact, this is being done today via agency measures of programs.  President Obama has directed agencies to identify their most important goals by July 30^th.   He says he’ll have his chief performance officer track and report to him on their progress.

This performance transparency can be to the public, or even just to government leaders.  The British government, under Tony Blair, created a “Delivery Unit” to track the key goals important to the Prime Minister.  The meetings of ministers that Barber organized to go over their agencies’ performance were closed to the public in order to ensure the discussions were candid.  However, the results were periodically reports publicly.  This kind of approach is being piloted in several U.S. state governments and cities. 

The leader of that unit, Michael Barber, said:  “Different public services will need different approaches to reform. The same services will need different approaches at different times.”  This was also the conclusion of several of the panelists at the Center for American Progress forum, as well. 

Tags: Bureau of Labor Statistics, Center for American Progress, CFO Act, Delivery Unit, McKinsey, Michael Barber, productivity, Reinventing Government, Tony Blair