Making environmental audits honest….Esther Duflo, Michael Greenstone, Rohini Pande, Nicholas Ryan
Reducing the auditors connection to the industries they are meant to check, and making them accountable to the public is key
Reducing the auditors connection to the industries they are meant to check, and making them accountable to the public is key
Speaking at the opening of an international seminar on “Global Environment and Disaster Management: Law and Society”, Chief Justice of India S.H. Kapadia highlighted a major problem with the procedure for environmental impact assessments of proposed industrial projects. Currently, industries select and hire the consultants who prepare the report needed for clearance. This system creates an obvious conflict of interest: the consultant, who is paid by the project proponent, has every incentive to deliver a report that is favourable to the industry.
External auditors more loyal to the firms under scrutiny than to the public interest are an old problem and not limited to the environmental sphere. In the US, the Enron scandal of 2001 exposed financial auditors who turned a blind eye to firms’ dubious accounting practices. Seven years later, the ratings agencies helped bring about the financial crisis by giving top marks to their clients’ lousy offerings.
Nevertheless, third-party auditing does have some advantages over traditional top-down bureaucratic regulation. Government regulators often lack the resources to keep pace with rapid growth in industry. This is why in 1996, the Gujarat high court ordered the Gujarat Pollution Control Board (GPCB) to set up a third-party environmental audit scheme, wherein industries from the most polluting sectors must retain an auditor to report pollution readings to the GPCB and to recommend improvements to industry practices. This scheme is in addition to the regular GPCB inspection system, which, due to insufficient manpower, does not inspect industries with the mandated frequency.
The scheme was designed with all the usual safeguards. Auditors must be certified, a given team can perform at most 15 audits per year and a firm can only audit the same industry three years in a row. But these safeguards did not relieve the fundamental tension highlighted by the chief justice, that consultants are still paid by, and therefore loyal to, the industries they audit.
To address this problem, the GPCB, with technical advice from the authors, implemented a pilot that modified the scheme on an experimental basis. Auditors were paid from a central pool, rather than directly by the firms, and randomly assigned to each industry, rather than being chosen by the industry. Back-checks involving an independent team from a local technical university, taking the same pollution measurements as the auditor shortly after the audit visit, were performed to ensure the validity of the audit report. Auditors’ payments were based partly on their accuracy, as measured by the back-checks. The original high court order permitted all these changes and had even specifically proposed back-checks, though they had fallen out of the scheme over time.
The difference in the reports submitted by the auditors working under the improved scheme, relative to under the traditional scheme, is striking. Auditors with incentives to be independent report significantly and consistently higher pollution readings.
Auditors working under the improved scheme were not more capable or otherwise different from those still in the traditional scheme — in fact, most of the audit firms were working under both schemes at the same time. Nor is the difference in reporting because the industries assigned in the improved scheme somehow pollute more. What did change, however, was the difference between the true reading measured by the back-checkers and the reading reported by the auditors. The same auditors were who provided low readings under the traditional scheme, often just below pollution control board norms, started reporting what they had actually measured when made properly independent.
The ministry of environment and forests is considering changes to the environment impact assessment procedure and several state pollution control boards are contemplating a greater use of third parties to help monitor industry and reduce pollution. This experiment from the Gujarat board shows that audits can work, but only if the scheme is designed so that the auditors’ loyalties lie not with the industries they audit, but with the greater public.
Duflo teaches economics at MIT and is director, Jameel Poverty Action Lab; Greenstone teaches economics at MIT;
Pande teaches public policy at Harvard’s Kennedy School and Ryan is a PhD candidate in economics at MIT. Greenstone, Pande and Ryan are working with the environment ministry on the emissions scheme