Behavioral Ethics

Credit: creative commons

Today the Behavioral Insights Team (BIT) from the UK released its annual report and (as anything released from the BIT) it was a delight to read it, with a small exception: Their partnership with Mexico. I am Mexican and I am a big supporter of applying behavioral insights to public policy, but this partnership makes me feel uneasy.

For those who may not know it the BIT is “ a social purpose company…dedicated to the application of behavioral sciences…(with the objectives of) making public services more cost-effective and easier for citizens to use; improving outcomes by introducing a more realistic model of human behavior to policy; and wherever possible,
enabling people to make ‘better choices for themselves’.(They achieve their objectives by) redesigning public services and drawing on ideas from the behavioral science literature” (source: The use of behavioral insights (a.k.a. nudges) has been largely criticized as a form of manipulation. A critic that is well founded but somehow irrelevant, in my opinion, as most people accept its use as long as the behavioral insights are applied to do well and are aligned with people’s interests, e.g. to promote health, safety, and clean energy.

So far so good. Now back to the BIT project in Mexico. Despite the stated objective of “increasing business formalization in Mexico” (a very laudable objective), the execution in partnership with the Mexican Ministry of Finance was simply to increase tax revenue from the business already in the formal sector. The BIT reports the procedure as “748,499 taxpaying businesses that had not submitted their declaration a fortnight before the deadline received either one of three SMS reminders or no reminder. All three reminders significantly increased declaration rates.”

The BIT reports that the project “increased declaration rates from 24 per cent to 33 per cent. The messages also continued to impact on declaration in the following four months with no further reminder. The increased revenues represented a return on investment of 400 per cent for the SMS, even though these firms currently still receive either a 90 per cent or 100 per cent discount on their payments.” This is great but it is not the main Mexico’s problem.

This is precisely what makes me uneasy. I do not feel comfortable at all with a project that helps a government to increase the revenue not by increasing the taxable base, i.e. increasing business formalization, but to collect more from those already in the narrow base. Specially when the informality economy in Mexico has been largely discussed and pointed as a main cause of inequality and poverty, reaching levels of up to 80% in some States (ref:

Another thought that makes me completely uneasy is that most probably the government will increase its revenues but this would not be of benefit for the people. Why? Simply because Mexico “is the poorest performer for safety and corruption across OECD countries”, is also the most corrupt within democratic Latin America, and according to the World Economic Forum’s corruption index (perception index), Mexico is the most corrupt country of the world. So, what are the odds of this money going to improve people’s life and not into one more corruption scandal? Pretty low odds.

Is it ethical that the BIT helps such a corrupt government to collect more taxes given the odds of misuse? Wouldn’t it make more sense that the BIT could help the Mexican government (and the Mexican people) to fight corruption FIRST, before increasing tax collection? It is well known in the behavioral field that “when you focus heavily on one thing, there is just less mind to devote to other things” which is known as “tunneling”. Would it be that BIT is falling prey of tunneling, so focused on proving the worth of applying behavioral insights that they just neglect the context? I have an idea to avoid this ethical problem and thus, tunneling in applying behavioral insights. If you happen to know somebody from the BIT, I will be so glad to share this idea with her/him, please let me know.