Enforcement and Deterrence with Certain Detection: An Experiment in Water Conservation Policy

The purpose of this study is to test the effectiveness of a new, automated outdoor watering enforcement system in reducing overall water use while minimizing the number of fines issued. The intervention is targeted to the universe of single-family households in Fresno, California with smart water meters.

2018-07-01

2018-09-30

The primary outcomes of the experiment can be grouped into three categories:
i. Hourly Water Use: This contains hourly water use mapped to a household ID for each household in the study.
ii. Enforcement Actions: This contains enforcement actions taken by the city, i.e. warnings and fines that have been issued by the city for non-compliance with the written ordinance. This information contains date of enforcement action and enforcement action type, and is linked to the household ID.
iii. Other City Actions: This contains information on other actions the City of Fresno has taken to foster water conservation, such as rebates that have been given to households for purchasing water-efficient appliances, free water timer tutorials that have been provided, and other initiatives including a customer support hotline.

Between July 1, 2018 and September 30,2018, all single-family residential households with smart water meters in the city of Fresno, California were randomly assigned to groups that vary along two dimensions: Enforcement Method and Fine Level.

Along the Enforcement Method axis, households were assigned to one of two groups: Visual Inspection (household receives a fine if a violation is detected by visual inspection–this was the system in place until 2017, or the control), or Automated Enforcement (household receives a fine if a pre-installed water meter measures hourly water usage above a given threshold–this is the treatment or “intervention”). The Automated Enforcement group is further randomized into three groups defining the amount of water utilization that triggers an enforcement action:
i. Baseline (300 gallons/hour threshold)
ii. Reduced Enforcement 1 (500 gallons/hour threshold)
iii. Reduced Enforcement 2 (700 gallons/hour threshold)
The Reduced Enforcement thresholds were chosen, respectively, to reduce the expected number of automated fines issued by 50% and 75% relative to the baseline threshold of 300 gal/hr.

Along the Fine Level axis, households were randomized to receive different fine amounts. Violators received a maximum of one fine per month. The municipal code outlines the following fine schedule:

i. First month in violation: $0
ii. Second Month in violation: $50
iii. Third Month in violation: $100
iv. Fourth month in violation: $200

All households were randomly assigned to be subject to one of three fine levels:
i. Baseline ($0, $50, $100, $200)
ii. 50% Fine Discount ($0, $25, $50, $100)
iii. 75% Fine Discount ($0, $12.50, $25, $50)
In total, all households were randomly assigned to one of 11 treatment groups or a control group.

Randomization was completed in an office using a computer.

Randomization was completed at the household-level.

No

89,581 households

89,581 households

- Control Group
1. Non-automated enforcement, Baseline fine group: 40,311
- Treatment Groups
Non-automated enforcement:
2. 50% fine level: 4,479
3. 25% fine level: 4,479
Automated: 300 gal/hr water use threshold:
4. Baseline fine: 4,479
5. 50% fine level: 4,479
6. 25% fine level: 4,479
Automated: 500gal/hr water use threshold:
7. Baseline fine: 4,479
8. 50% fine level: 4,479
9. 25% fine level: 4,479
Automated: 500gal/hr water use threshold:
10. Baseline fine: 4,479
11. 50% fine level: 4,480
12. 25% fine level: 4,479

New technologies allow perfect detection of environmental violations at near-zero marginal cost, but take-up is low. We conducted a field experiment to evaluate enforcement of water conservation rules with smart meters in Fresno, CA. Households were randomly assigned combinations of enforcement method (automated or in-person inspections) and fines. Automated enforcement increased households' punishment rates from 0.1 to 14%, decreased summer water use by 3%, and reduced violations by 17%, while higher fine levels had little effect. However, automated enforcement also increased customer complaints by 1,102%, ultimately causing its cancellation and highlighting that political considerations limit technological solutions to enforcement challenges.

Browne, Oliver R., Ludovica Gazze, Michael Greenstone, and Olga Rostapshova. "Man vs. machine: Technological promise and political limits of automated regulation enforcement." The Review of Economics and Statistics (2023): 1-36.

https://doi.org/10.1162/rest_a_01316