Climate Change, Muni Finance and Machine Learning

Trying to create the business case without numbers

A fundamental challenge to creating the business case for climate resilience is the inability to readily access and aggregate financial data.  This is especially true at the municipal level.  Although municipalities amass a dizzying array of financial data, the way in which it is reported precludes real-time analytics or large-scale, aggregated comparisons across municipalities.  It also fails to capture or disclose the actual costs of climate change.

Why does this matter?

Putting climate aside for a moment, the inability to access real-time muni finance data means that the reporting winds up to be historic in nature. It is comparable to wanting to get a reading of a patient’s current blood pressure but not having access to that data – or the rest of their health screen – until a year later, potentially too late for a preventative intervention.

In addition, current muni reporting often misses key trends or inflection points since information is often averaged across a year’s worth of data.  This has the potential to effectively mute or cloak some otherwise significant signals.   For example, a recent study has shown that traditional public fiscal health indicators can fail to show the actual costs that were incurred during some extreme weather events.  Without being able to easily extract that information, the economic relevance of these costs and the political will to address them can also be lost.


Hurricane Laura flooding near Lake Charles, LA (Source: Orlando Sentinel/AP Photo/David J. Phillip, accessed 1/4/20).

Even once the annual financial results have been prepared for reporting, they are not often certified or widely-distributed until several months after the close of the reporting year.  This means that a good chunk of the reported information is – at best – three to six months out of date, with some being as much as 12-18 months out of date. 

And, finally, the majority of the reports themselves are generated as pdfs that are NOT machine-readable.  That means that any large-scale analysis of data requires additional months (or years) of tedious work to pull together what is already outdated information.  Instead of spending time analyzing the data to better inform municipal strategies, analysts and municipal officials alike are forced to continue to drive in the dark without headlights and navigating the roadway largely with their rearview mirror.

Using this current model, how can municipalities ever hope to come up with meaningful or transformational solutions (something that climate change will require) when sensitive periods for action have already passed – just because we haven’t developed the reporting tools to allow for forward-looking policy choices and analytics? 

Luckily, there is a movement afoot to change that. And this is where machine learning comes in.


Finding the numbers

The concept of creating a tool for electronic financial reporting was introduced in 1998 and eventually developed into the eXtensible Business Reporting Language, or XBRL as it is currently known today. Both XBRL and Deloitte offer efficient overviews of this concept, and my colleague, Liz Sweeney of Nutshell Associates, provides a very informative deep dive of the subject as it relates to municipalities as part of the Municipal Beat podcast.


Overview of how the current data collection system (used for reporting purposes) currently operates in most US municipalities (Source: XBRL).

In short, XBRL is a tool in which entities can electronically “tag” key aspects of their data that allow them to be easily read by machines and uploaded for further analysis.  It is a free, open software – one that has been widely used by corporations and which is now being heavily promoted for use within the public sector.  It has also been gaining traction with the Sustainability Accounting Standards Board (SASB) and others for the standardized reporting of material EGS metrics.

While this process does not completely solve for the real-time reporting needs, it does immediately alleviate some of the reporting delays detailed above.  It also brings a greater internal efficiency to the municipalities since they will be able to readily reuse, repackage and report data to others – and even to meet their own internal operational needs – without having to reinvent processes for each.  Additionally, it moves us closer to “one source of electronic truth” which will allow for more timely and relevant large-scale regional and national assessments.

There is reason to believe that as these processes become more streamlined, it will also become less cumbersome to introduce other metrics and indicators – including those which have a greater relevance to climate risk.  For example, the ability to capture the actual costs of extreme weather events BEFORE they are reimbursed; the ability to track cumulative impacts of shortened life expectancies within infrastructure as a result of climate stress; tracking development and economic implications within municipalities and regionally; climate migration trends, etc. 

All of these data points are essential information as we create the business case for climate resilience.  The existence of a readily-accessible, machine-readable, “one-stop-source” of information for municipal finance is an important step in helping communities both understand and proactively prepare for climate change.  It will allow us to further highlight – with financial quantification – the interdependencies between climate resilient communities and economies.r

Get in Touch

If your project is a good fit, we’ll reach out to schedule a complimentary discovery call for next steps.