Predictive Value

Predicting violations – The case of the outlying ratio

In the last blog we looked at different industries and how they deal with mitigating environmental impact. We used the waste-to-disposal ratio to assess an industry’s efforts and success in reducing the release of regulated chemicals into the air and water.

The same measure can be used to predict whether a company is likely to get into trouble with the regulator and be cited for violations of the Clear Air and Clean Water Acts.

In this example of using the data, we took three types of waste chemicals that are commonly used in manufacturing and are prevalent in many industries: Ammonia, Aluminum and Zinc. We calculated the waste to disposal ratio by company for each of these three chemicals and then looked at the distribution of the results.

In most cases, the calculated values cluster around a stable value – an increase of 1 pound in waste production leads to a proportional increase in disposal. The ratio may not be exactly 1. As we have seen in the previous blog, some industries, and therefore companies, are better at capturing and re-using or neutralizing their waste. For those companies, an increase of 1 pound in waste will lead to an increase of less than 1 pound in disposal.

Then there are those companies where the ratio is much too low compared with their peer groups. We have found that frequently report outliers are significantly more likely to be found in violation of environmental laws. The pattern is clear enough that it suggests causation. It looks like regulatory agencies, typically on the state level, are using the outliers to target companies for inspection,

Low disposal-to-waste ratios versus violations

A look at the permits and violations data by facility confirms this suspicion. Facilities with erratic waste-to-disposal ratios that lie outside their industry norm are more likely to be inspected than those reporting stable ratios that are close to the numbers filed by their peers.

Inspections versus violations for ammonia emissions

The emission data can therefore be used to predict which companies are more likely to be targeted by regulators, be found in violation, and be assessed fines. The penalty amounts are too low to have an impact on any but the smallest operators, but companies may suffer adverse effects on their reputation if often found in violation.