The Census Bureau's Governments Division collects financial data on the nation's state governments, its more than 89,000 local governments and its more than 5,500 public pension plans, as well as government employment and organizational data. Due to the collapse of congressional budget processes and downward pressure on non-defense discretionary spending, the division receives limited funding, and even that is periodically disrupted by government shutdowns. The result is predictable: decreasing levels of service and a lack of investment in new technologies that could lower the costs of collecting data while improving its quality.
In the private sector, innovations such as cloud computing and machine learning are enabling companies to use fewer people to collect more data at higher quality. But while Census Bureau employees take advantage of new technologies when they can, budget constraints limit the Governments Division's ability to fully embrace them.
This is unfortunate, given increased demand for government finance information. Since the Great Recession, many state and local budgets have been under stress. Some jurisdictions have experienced genuine fiscal crises. This has increased worries among creditors and state overseers about government financial sustainability. These stakeholders need more government finance information, and they need it more quickly than before, as do academics and think-tank analysts monitoring public pensions, Medicaid spending and other fiscal pressure points.
The Census Bureau is unable to keep up with these increasing needs. Statistics often appear two years after governments' fiscal years end. While aggregate data published by Census is reasonably accurate, data for individual governments -- especially smaller ones -- is subject to significant error. Finally, the data elements the bureau reports for any given entity are insufficient for municipal credit analysis.
Market data providers, academic researchers and think tanks have filled some of the gaps in Census data with a hodgepodge of reports and services. In many cases, these alternatives rely on data extracted from Comprehensive Annual Financial Reports filed by state and local governments. Private data-collection efforts are often duplicative and typically have limited coverage.
The public interest in timely, comprehensive and accurate state and local finance data would be better served if private data collectors and technology providers could somehow join forces with the Census Bureau in a P3. Of course, government agencies contract out services to private companies all the time, but a data-collection P3 could work differently.
A foundation, for example, might pay a private company to develop new data-collection technology and then contribute it to the Census Bureau. Or companies, academic departments or nonprofits might collect data on behalf of the bureau, financing their work with subscription revenues. Subscribers could get enriched or more timely data for the fees they pay.
A first step toward a data collection P3 would be the development of data standards that can meet the needs of both the Census Bureau and non-governmental participants. These standards should be based on XBRL (eXtensible Business Reporting Language, which is used by corporations for the quarterly and annual financial statements they must submit to the Securities and Exchange Commission).
There are already signs of progress on this front. XBRL US, the standards body that manages XBRL domestically, has set up a State and Local Disclosure Modernization Working Group that includes nonprofits, academics, companies and governmental bodies, including the Census Bureau. The working group's main task is to create a reporting taxonomy consisting of financial data elements required by the Census, municipal finance analysts and researchers.
Once this taxonomy is in place, we can build data-collection and reporting tools that will reduce the Government Division's workload while producing more timely, comprehensive and accurate state and local government finance data - at no additional cost to the taxpayer.
