Content Navigator

Is Process Mining Right for My Organization?

Is Process Mining Right for My Organization?

Process mining is an exciting and potentially useful set of tools, but only one of many worth considering. Even if process mining is right for your organization, there is a better place to start. Use your data and economic principles to gain an understanding of the relative importance of different opportunities to create value. That way, you can deploy your organization’s scarce talent and other resources more optimally.

Innovate - Spring 2021 Digest

Innovate - Spring 2021 Digest

This quarter, Innovate focuses on why a data strategy can be so valuable. We also turn to the pandemic and the question of how its economic impact varies over time, industry, and location, presenting a public dashboard on this topic. Even as the pandemic subsides in the US, it rages on elsewhere. Our thoughts are with our colleagues, clients, and the broader community whose families continue to be severely affected by this crisis.

Building a Data Warehouse? Consider Marie Kondo's Approach Instead

Building a Data Warehouse?  Consider Marie Kondo's Approach Instead

If your current data infrastructure doesn’t “spark joy,” read this explainer before investing in an Enterprise Data Warehouse. We argue that firms should think about business objectives, analytics, required data flows, and tools – in that order – and that modern tools and services offer potentially big benefits over a traditional data warehouse given management’s goals for the business.

A New Metric for COVID’s Evolving Economic Impact

The Daily COVID Economic Impact Index (EII)

The country’s uneven recovery from the pandemic means that COVID-induced economic volatility will continue to be a major decision-making factor. Tracking leading indicators of COVID’s economic impact can help investors and managers better anticipate the impact of the pandemic on their growth plans and prepare employees and customers for what lies ahead. As other researchers have shown, county-level variation in COVID patterns can have important economic implications (e.g., Noland and Zhang 2021).

We developed an index, the Daily COVID Economic Impact Index (EII), to help throw more light on possible patterns in the current economic impact of the ongoing pandemic. This index attempts to capture the substantial variation in COVID’s economic impact across locations due not only to day-to-day differences in new case counts across counties, but also to differences in the concentration of employment across industries — differences that lead some counties to be more economically sensitive to COVID than others. Specifically:

  • Industries vary in their sensitivity to COVID. Nationally between January and April 2020, aggregate earnings fell 52% in Leisure and Hospitality (L&H) vs. only 1.1% for Financial Activities.

  • Locations vary in their economic structures or orientations toward COVID-sensitive industries. For example, in Mono County, California, L&H employs more than half (58%) of workers vs. only 5.6% of workers in Collier County, Florida.

  • Hence, the overall economic impact of COVID varies by location according to the location’s economic structure, meaning that given number of new cases generally won’t have same economic impact across two or more locations.

The index is normalized so that counties with the same employment-industry orientation as the US overall (e.g., San Bernadino, CA and Osceola, MI) are assigned an EII value equal to the unadjusted count of new COVID cases per day capita. Counties with higher employment shares in COVID-sensitive industries (those with economic sensitivity factors >1.0) have an EII greater than the unadjusted case count. Counties with lower employment shares in COVID-sensitive industries (those with economic sensitivity factors <1.0) have an EII less than the unadjusted case count, reflecting their relative insulation from the economic impacts of COVID.

We present this new index in the dashboard linked below. The dashboard contains a map and other visuals showing the distribution of the EII across counties, with filters that allow users to select states and counties of particular interest. Results are updated daily to reflect new case counts as reported to The New York Times.

More below on how managers and investors can derive value from this tool.

Dashboard updated Daily.

In April 2021, the map of the EII showed “economic hotspots” that paralleled COVID hotspots at the time (higher intensity in various MI and TX counties, lower intensity in select CA counties). The map continues to evolve with the spread of the Delta variant and beyond. Individual organizations can identify what the economic impact looks like for their own customer and employee footprints. On the bubble chart, we can see there are some larger counties (e.g., Clark, NV and Orange, FL) that have relatively high EII values compared with other counties (e.g., Fairfax, VA; Fulton, GA; Middlesex, MA) with similar daily COVID case counts — indicating that COVID may currently pack a more potent economic punch in these counties (i.e., the NV and FL counties compared with the VA, GA, and MA counties).

The Dashboard

How to use The Dashboard

1. Map

This economic impact map plots the EII for each county, allowing viewers to visualize the potential economic footprint of new COVID case counts as it evolves day-to-day.

The filters at the top of the page allow viewers to zero-in on selected states.

Many smaller counties report zero new cases in any given day, resulting in an EII of zero. Counties with an EII of 10.0 or higher are shaded progressively more red, with counties above EII=75.0 shaded darkest red. We chose these color cutoffs based on the current distribution of the EII across counties and may update them in the future.

2. New cases vs. EII

Viewers can use this bubble plot to quickly see which counties are more and less economically sensitive to COVID according to our index. The plot shows the relationship between the raw/unadjusted count of new cases and the EII across counties. The size of each bubble reflects the county’s population. Filters at the top of the page allow viewers to zero-in on selected states or counties or visualize only the most populous counties.

The EII is normalized so that counties with the same employment-industry orientation as the US overall (e.g., San Bernadino, CA and Osceola, MI) have an EII equal to the unadjusted count of new COVID cases. In the bubble chart, such counties lie on the 45-degree line. Counties with higher employment shares in COVID-sensitive industries (shaded in red) will have an EII greater than the unadjusted case count and thus lie above the 45-degree line. Counties with lower employment shares in COVID-sensitive industries (shaded in blue — these are counties that are more COVID-insulated than the country as a whole, from one economic standpoint) will have an EII less than the unadjusted case count and thus lie below the 45-degree line.

3. Top Counties

The two tables on this page display the counties with the largest and smallest EII values on any given day. Additional descriptive statistics are provided at the bottom of the page for context on the extent to which these counties are outliers.

4. Distribution across Counties

This histogram provides additional context for how the EII is distributed across counties on any given day. Viewers can also click on a bin to see a list of the counties with an EII in the selected range appear in the table.

Rationale for our Approach

Why calculate an index for the pandemic’s evolving economic impact rather than look directly at what’s happening with each county’s economy? Our objective is to provide insight into the potential ongoing economic impacts of COVID. But in addition to COVID, a lot of other factors are likely to affect a county’s economy. Isolating individual causal factors is very difficult. To do this, we designed an index to focus in on the pandemic’s uneven impact on the labor market according to industries — a narrower aspect of the economy that is particularly likely to be influenced by COVID in a way that our index captures.

Underpinning our index is the idea that the pandemic’s economic impact is a product of the pandemic’s intensity at a given point in time and the sensitivity of a county’s economy to a given intensity level. We capture intensity of the pandemic for each county by measuring the number of new cases per capita using data reported by The New York Times. We capture each county’s economic sensitivity by combining two pre-pandemic variables - the distribution of employment across major industries in the county, and the shock to aggregate labor market earnings in each industry that happened nationally during January-April 2020 as the pandemic took hold of the economy. We obtained these data points from the US Bureau of Labor Statistics and the US Census Bureau.

Our index arguably captures some of the causal impact of the pandemic on the economy because it focuses in on a more specific, plausible channel of causation (different employment shares by industry that cause counties to have different economic sensitivities to the pandemic) that was arguably in place prior to the pandemic (i.e., our hypothesized channel of causation wasn’t itself caused by the pandemic). That said, it is far from perfect. Industries and counties have now adapted to the pandemic, so a new case doesn't affect things same way it initially did. The economy won't follow same path back during recovery that it went down during the onset of the pandemic. The relationship between the pandemic’s intensity, a county’s economic sensitivity, and the impact of the pandemic on the economy does not take the same exact shape that our index assumes (the actual shape is unknown and likely nonlinear in numerous factors). And so our index is at best indicative of direction of relative economic impacts of the pandemic.