Misaligned incentives: how certification schemes can evolve to improve quality

 
AAMAAQDGAAoAAQAAAAAAAA9jAAAAJDJmZDNhMzQxLTRhZTItNDg3Ni1hYzYxLTljNWZkYjlkZmIyYg.png

My recent article on the limits of certification in agriculture prompted some interesting discussions on the future of this tool and how it can improve, from a compliance point of view. This post is the first of several from me and my colleagues at SAN on how certification can evolve, based on our experience with different schemes over the years. The series is meant to support further reflections on how to improve these initiatives.

While some of the limitations of certification are structural, there are other points on which it can evolve. One challenge that is underappreciated is certification systems that adopt business models that have misaligned incentives between the scheme owners and certification bodies (CB).

Most certification initiatives operate through multiple CBs that compete with one another for the provision of a commoditized service, the certificate of that scheme. The certificate issued by one CB has the exact same value as the certificate of another CB. Quality assurance is done through each CB’s internal quality control system and usually by an accreditation body, typically with a strong focus on processes. These constitute weak feedback loops to promote quality.

In practice, competition between the CBs is primarily driven by price, which can lead to the reduction of time spent in the field and the use of less experienced auditors and certification decision makers, resulting in poorer quality in implementation. The fact that many producers get certified simply to comply with buyers’ requirements and not out of a perception of value further reinforces the downward pressure on CB service costs.

In addition, the issue of CB conflict of interest needs to be better understood. Too often the response has been “that is just how things are”. There is already evidence[1][2] that indicates that auditing bodies arrive at significantly different conclusions if they are paid by the subject of their audit or by a central organization or entity further downstream. Simply put, CBs may see certified operations as their clients and not the certification scheme, and the need for client acquisition and retention may consciously or unconsciously bias their work.

The result of the above is that the CB’s internal quality control systems and accreditation body are always swimming upstream, engaging after the work on the ground has been done and pushing quality against the natural dynamic of the model, which is to lower service costs. This design can contribute to faster growth for the scheme, but it is not likely to consistently deliver quality.

What are possible design changes to improve this issue? Each alternative will have trade-offs, but here are some examples of designs that are better suited to deliver quality:

  • Limiting the number of CBs that operate in given region and exercise stronger a priori oversight on their proposals and on how certificates get transferred between CBs. To a large extent, this was the model adopted by SAN when it was working with certification. It brought some added transaction costs and worked better in some regions than in others. 
  • Assigning the responsibility over a given region to one single CB. Considering that monopolies tend to generate inefficiencies and stifle innovation, controlling service costs and periodically reviewing this assignment also becomes very important.
  • Allocating each case to a CB, based on the quality of technical proposals and costs. This presupposes a centralized function, which would add transaction costs to the scheme. Transparency in the allocation of cases would be essential for such a model to work.
  • Incorporating feedback mechanisms for workers, their organizations, civil society and other stakeholders that can influence the scheme’s quality assurance and decision on choice of CB.
  • Change to a model where audits are not paid by the producers.

The above is not a comprehensive list nor necessarily mutually exclusive one, and each option will have trade-offs between quality and costs. While there are other actions that can be taken to improve the quality of certification, such as higher use of surprise audits, focus on auditor performance and rotation of auditors, ensuring that there is better alignment between the incentives for the scheme owner and the CBs, its implementation partners, is probably one of the most important changes that can made by certification initiatives to safeguard and improve their credibility.

[1] http://www.hbs.edu/faculty/Publication%20Files/14-032_a106325c-070b-403d-8d60-9d037bcc3162.pdf

[2]https://cepr.org/sites/default/files/The%20Quarterly%20Journal%20of%20Economics-2013-Duflo-1499-545.pdf

 
Andre de Freitas