When Performance Measures Drive Bad Behaviour: Lessons for Regulators

The Wells Fargo Cross-Selling Scandal

A great piece in this week’s Harvard Business Review (“Don’t Let Metrics Undermine Your Business”) by Michael Harris and Bill Taylor, about the problem of performance metrics. While the article is aimed at private sector companies, the analysis applies equally well to the regulatory sector.

Regulators have been working hard in recent years to develop performance measures. This is a laudable task, but my concerns parallel those raised by the authors in the HBR piece.

Due to the human brain’s tendency to surrogation, staff will substitute easy-to-grasp concrete performance measures (e.g., time to complete an investigation) for more abstract strategy. To some degree, this makes sense. We develop strategies, and try to measure our performance against them. In our minds, the metrics and the strategy are one and the same. But they are not.

We know that performance measures are always imperfect; to become obsessed by them is a mistake, particularly where the measures are in some ways perverse, and take place within the context of a bad culture.

The authors use the recent cross-selling scandal at Wells Fargo as an example. Sales staff were given high (and arbitrary) sales targets to meet, to cross-sell clients on multiple bank products. Due to deep and pernicious cultural issues, including tolerance for questionable sales practices, a relentless focus on numbers, and pressure from managers whose incentives depended on meeting sales quotas, sales staff began opening accounts and issuing credit and debit cards without customer authorization. After multiple investigations and hearings, the CEO resigned, the leadership structure was changed, and the bank paid over a billion dollars to financial regulators as settlement.

According to this Harvard Law School Corporate Governance and Financial Regulation analysis of the scandal:

Some outside observers alleged that the bank’s practice of setting daily sales targets put excessive pressure on employees. Branch managers were assigned quotas for the number and types of products sold. If the branch did not hit its targets, the shortfall was added to the next day’s goals. Branch employees were provided financial incentive to meet cross-sell and customer-service targets, with personal bankers receiving bonuses up to 15 to 20 percent of their salary and tellers receiving up to 3 percent

Of note, Wells Fargo was once known for sound banking practices, and emerged as the least-tarnished of the big banks following the 2008 financial crisis. Yet within a matter of years, something changed. A key plank in Wells Fargo’s strategy – deep customer relations – became entirely subverted by the bad metric, combined with bad culture and a perverse incentive culture.

Lessons for Regulators

What is the moral of the story for regulators grappling with the effort to measure their performance in some meaningful way?

I will summarize the lessons for regulators as follows:

1. Involve key management in strategic planning. Do not set strategy in an executive vacuum; otherwise it will always be a confusing, meaningless abstraction for managers and employees.
2. When planning strategy, focus on internal measures in addition to external measures (e.g., stakeholder awareness), so that strategic performance measures are not entirely irrelevant to operational staff.
3. In developing performance measures, be transparent about the limitations of such superficial measures as turnaround time. In the regulatory world, timelines are important for reasons of statutory compliance, fairness, and stakeholder satisfaction. But they never tell the full story. For example, if managers pressure investigative staff to complete investigations within a given (and perhaps arbitrary or unrealistic) time frame, understand that you may be motivating staff to game the system, to cut corners, to skip important risk assessment, investigative planning and execution steps, and to skimp on procedural fairness.
4. An excessive emphasis on timelines can also lead to a kind of rigid process blindness, in which staff are unable to effectively deal with novel information that should affect risk assessment, or other requirements (such as accommodation requirements under human rights legislation), due to their fear of missing timelines.
5. While it’s necessary to measure regulatory program performance activitiesand outputs (or deliverables), regulators should focus performance measurement where possible on outcomes, and ensure that the designation of these outcomes results from deliberate analysis by staff and management:
6. For example, investigations departments should consider supplementing timeline metrics with metrics reflecting quality and fairness of investigations and regulatory decision making.
7. Regulatory managers may not be driven by bonuses in the same way their private-sector counterparts are (although regulatory executives are increasingly being given performance incentives). But it is essential to be aware of the cultural example management is setting. If poor-quality work, corner-cutting or unfair processes are tolerated in order to meet timelines, the accountability for that deficient work product rests with management.