OpenGov Voices: A Transparent Approach to Understanding Local Government Debt


Disclaimer: The opinions expressed by the guest blogger and those providing comments are theirs alone and do not reflect the opinions of the Sunlight Foundation or any employee thereof. Sunlight Joffe_Headshot_1Foundation is not responsible for the accuracy of any of the information within the guest blog.

Marc Joffe is the founder of Public Sector Credit Solutions (PSCS) which applies open data and analytics to rating government bonds. Before starting PSCS, Marc was a Senior Director at Moody’s Analytics. You can contact him at Marc is also one of the winners of Sunlight Foundation’s OpenGov Grants.

High profile bankruptcy filings by Detroit and other cities, along with concerns about public employee pensions, are increasing borrowing costs for state and local governments. Higher interest payments to bondholders mean higher taxes and fewer services. However, with transparent data and analytics, local government bonds can get reasonable interest rates — as this post will illustrate.

Over the last 70 years, municipal bond defaults have been rare. In a typical year, no more than one in 1,000 municipalities fail to make timely payments on their tax supported debt. Also, interest on municipal bonds is exempt from federal income taxes and usually free of state income taxes as well.

Because of their low risk and favorable tax treatment, municipal bonds have typically yielded less than US Treasury bonds – making it easy for states, cities, counties and school districts to finance new infrastructure. Time series data available from the Federal Reserve (a portion of which is depicted in the accompanying graph) show that yields on “munis” were lower than Treasuries from 1953 until the 2008 financial crisis. This discount returned briefly in 2010, but since Meredith Whitney predicted a wave of municipal bond defaults on 60 Minutes in December 2010, muni yields have exceeded Treasury yields – often by substantial margins.

Municipal and Treasury data

Whitney’s forecast of 50-100 sizeable defaults within a year failed to materialize, but the few bankruptcies that have occurred (most notably Stockton and San Bernardino, California in addition to Detroit), have continued to incite fears of a meltdown. Although companies default on their bonds far more frequently than governments, cases of municipal fiscal distress receive far more media attention, are much more likely to become politicized, and thus tend to reinforce an apocalyptic view of municipal finance.

Rating Agencies Not Coming to the Rescue

Market mechanisms often correct financial system imbalances – but they seem to be failing in this case. Bond investors have historically relied upon credit rating agencies to help them make sense of debt securities, but these agencies lost credibility during the financial crisis. This loss of trust was triggered by their assignment of artificially high ratings to mortgage backed securities during the housing bubble. These supposedly safe bonds suffered heavy losses during the crisis. Earlier this year, the Department of Justice sued one of the rating agencies, Standard & Poor’s, for its role in the financial crisis.

Ratings inflation on mortgage backed securities is part of a much larger problem that has compromised the authority of rating agencies and harmed local governments. The problem is that rating agencies assign similar letter grades to different types of bonds, but these letters have inconsistent meanings in terms of credit risk. Municipal bonds receive the harshest treatment. Although Dodd-Frank Section 938 technically forbids this practice, rating agencies circumvent or simply ignore this legislation.

While most people familiar with financial markets know that there were serious problems with ratings on residential mortgage backed securities, fewer know about the inconsistencies across other asset classes such as municipal bonds and commercial mortgage backed securities. A simple example should help demonstrate this inconsistency.

Recently, Moody’s assigned its highest rating (triple-A) to bonds backed by a mortgage on a single shopping mall in northern Virginia—Tyson’s Galleria. If this mall is destroyed in a natural disaster, or if it goes bankrupt due to competition, investors in these triple-A bonds will lose money. The latter scenario recently occurred to the Tri-County Mall in Cincinnati, inflicting losses on triple-A investors in its mortgage bonds.

Meanwhile, Moody’s rates the general obligation bonds issued by the state of California A1 – four steps lower on the agency’s 21 step scale. This, despite the fact that no state has defaulted on general obligation bonds in 80 years, and that California now has a surplus.

The Solution: Transparent Data and Analytics

So if today’s rating agencies are confusing investors rather than informing them, how can we solve the problem of informing municipal bondholders and thereby achieving more reasonable interest rates for local government bonds? The answer, I believe, lies in making more effective use of publicly available government financial data.

Since the 1960s, economists have been creating models that estimate corporate bankruptcy risk by comparing financial statements of companies that go bankrupt and with statements from those that remain solvent. Statistical analysis of these statements yields an equation that estimates the probability of default. Once the equation has been derived, new financial statement data can be plugged into the model to obtain default probability estimates. An investor can then use these default probability estimates to determine the fair yield on bonds issued by these companies – a fair yield being one that compensates her/him for the bond’s default risk.

Since most governments are required to produce audited financial statements and make them publicly available, default probability models and estimates for governments can be created with open data. This is the goal of the California Local Government Credit Scoring project, recently funded by a Sunlight Foundation OpenGov grant.

Using the proceeds from an earlier grant, we created a basic default probability model and ran it against financial statistics for 260 California cities with over 25,000 people. One of the biggest challenges was to gather financial statements and extract relevant statistics, as I discussed in an earlier Sunlight blog post.

We will be using the OpenGov grant to extend coverage to California counties. We look forward to collaborating with other local transparency groups, think tanks, academics and financial market participants to expand coverage beyond California, add updated data for existing governments as it becomes available and improve the default probability model.

Once a tool like ours’ contains a critical mass of data and credible insight, investors will be able to trust it to distinguish the relatively small number of risky municipal bond issuers from the much larger universe of safe cities and counties. Greater transparency around municipal finance should produce lower interest rates and thus greater value for taxpayers.

Interested in writing a guest blog for Sunlight? Email us at