Overview
The Industry Cluster Identification report helps you explore how industries group together into clusters that better represent the structure of a regional economy. It is available in the Industry module of US Developer.
Unlike looking at industries in isolation, clusters highlight the interconnections among industries that share supply chains, technologies, labor needs, or demand drivers.
Clusters are especially useful for economic development, workforce planning, and education alignment because they reveal concentrations of related industries that form a region’s competitive advantages.
Data Sources
This report draws on Lightcast’s core US industry data, which integrates federal datasets such as the Bureau of Labor Statistics’ Quarterly Census of Employment and Wages (QCEW) and the Census Bureau’s County Business Patterns (CBP). Lightcast unsuppresses and enhances these inputs to provide full coverage of employment and earnings across industries.
The clusters in this report are based on the industry cluster definitions developed at the Institute for Strategy and Competitiveness at Harvard Business School (Michael Porter and colleagues). These definitions are widely used in academic and applied economic development research.
Lightcast Modifications to Harvard Clusters
When Lightcast (then Emsi) first implemented Harvard’s cluster definitions in 2014, we made several adjustments:
Inclusion of excluded industries: Harvard’s original implementation relied on County Business Patterns, which omits some industries (such as government and crop production). Emsi added these back in. For example, Crop Production and Animal Production & Aquaculture were added to the Agricultural Inputs & Services cluster, and Rail Transportation and Postal Service were added to Transportation & Logistics.
Expansion of existing clusters: Several financial industries (like Pension Funds and Health & Welfare Funds) were added into the Insurance Services cluster.
Creation of new clusters: Because government services were entirely absent from Harvard’s framework, Emsi created new clusters for Federal Government Services, State Government Services, and Local Government Services, each containing the appropriate NAICS industries.
These modifications ensure that all NAICS industries are represented in the cluster framework, including those essential for understanding the structure of regional economies.
Updating with NAICS
The original Harvard clusters were built on NAICS 2012. Since then, Lightcast has maintained the framework by remapping it forward to NAICS 2017 and then to NAICS 2022. This ensures that clusters remain compatible with the most current industry data. The “Industry Cluster Identification” report in Analyst today reflects Harvard’s framework, Emsi’s 2014 modifications, and Lightcast’s ongoing remapping to NAICS 2022.
Scoring and Ranking
Each industry in the region is evaluated in terms of 5 performance metrics: Earnings, Growth, Regional Competitiveness, Regional Specialization, and GRP.
Each industry receives a score of 1-100 for each performance metric using a process called rescaling. During rescaling, the lowest performing industry is assigned a score of 1 and the highest is assigned a score of 100. Other industries are then scored according to their distribution between the industries with the highest and lowest values.
If a weight has been applied to a certain performance metric, it will multiply all values of that metric. For example, if an industry received a growth score of 68, and growth had received a weight of 4x, that score would be multiplied by 4, yielding a score of 272. This effectively gives a weighted metric greater influence on the overall score for that industry.
Each industry's weighted scores by performance metric are totaled into an overall score, and then industries are rescaled again to determine overall score on a 1-100 scale.
Clusters are scored by the average score of their component industries. The averages are weighted based on employment, so industries with no employment will have zero impact on cluster scores, while the impact of larger industries will be proportionate to their size.