The Corona virus renders visible the dependence work plays in our society. With both the near-future threat of robotization, and the now-and-later threat of viruses, it is time Mexico adopts an unemployment insurance program. This essay outlines my own plan for this policy.
The fourth transformation initiated or restarted various social programs that didn’t exist in a serious form before 2019. These include amongst many others universal pensions; support for working mothers; income support for disabled people; conditional cash transfers for school going youth, employment programs for young adults, etc. Our current global pandemic highlights the importance of such income support and replacing measures, as they allow people to stay indoors. But it also revealed a gap: there are very little measures protecting working adults (especially men) from prolonged unemployment. Such a program only exists in Mexico City, but is lacking for the general population. In what follows I will first explain what unemployment insurance is, before outlining how this factors into our two most immediate threats: virus outbreaks and robotization. Finally, I will present a dynamic model for unemployment insurance that my team developed over the past years.
What is unemployment insurance? It a social policy that takes the form of periodic payments that support workers who have recently lost their employment involuntarily. It is different from severance pay in so far that a) it is monthly, b) it is not paid by the employer, but by the program. In most cases, people need to first ‘built up’ by having worked for a while, and need to accept any requirements the government asks of them (for example taking a course). Unemployment insurance helps to: a) protects workers from poverty, b) provide macroeconomic stability due to demand smoothening; c) improve labor markets by directing people towards the right jobs. This is common in almost all countries in the industrialized world, including many Latin countries: Brazil, Argentina, Chile, Uruguay, Venezuela and Ecuador. There is no true international standard, and each country has its own way of organizing this, with variations in duration, compensation, requirements, funding, etc. This means that there is a lot of room for adapting this principle to political preferences and national needs and capacities. Yet besides a failed attempt at implementation by EPN in 2013, Mexico never took the step in protecting workers from structural unemployment.
Besides that it is an explicit Human Right (Article 25), I will elaborate on two distinct reasons why Mexico should urgently adapt my proposal: pandemics and robots.
My proposal probably comes too late for the current Corona crisis, but we should remember that the last H1N1 outbreak was only a decade ago. It is thus not unreasonable to assume such a situation will repeat itself in the near future. How is Unemployment Insurance relevant here? The main form of combating viruses once they have entered the population is social distancing (and hygiene). This is very difficult in the workplace, which is why many people work from home if they are able to. However for the working class majority of this country this is not an option. Just under half the Mexican population lives in income poverty, and earns less than 3200 pesos (in Cities) and 2100 pesos (on the countryside) per month. For these families, not working equals perishing. If elites want (or force) people to stop working, we need an income compensation mechanism that allows them to ‘land’ without dying of hunger. This is both a matter of public health and social justice. It also protects those that involuntarily are out of work due to crisis layoffs or unpaid temporal closures. Finally, we must keep in mind that school closures put further financial pressure on families that don’t have domestic support, as it reduces the capacity to generate income. Even many middle class families would find relief in a modest income support that allows them to survive without losing their social position. Since everybody can get the virus or work at a company that closes, covering all workers makes sense.
The massive replacement of human workers by robots and artificial intelligence in the near futures is a second reason why Mexico should not wait with adopting unemployment policies.
In a previous essay I wrote about how robotization will rival climate change as one of the most fundamental challenges to our social contract of our lifetime. While the coming crisis might interrupt or accelerate it, this wave is coming and government responses across the world have been slow or absent. Roughly half the existing jobs could disappear, leading to unseen levels of unemployment. In contrast to previous industrial revolutions, this would have a much more broader impact on the population, which creates the solidarity to implement unemployment insurance. A street survey of 366 people in Nuevo León conducted by my own team shows that 86% of people consider their job at least partly replaceable by robots. In our conservative state, 86,9% of respondents support the principle of unemployment insurance as a response to automation, and 91,6% thinks this should cover all workers, not just those affected by automation! This offers a new window of opportunity to finally implement this policy and catch up with the rest of the world. Of course, unemployment insurance by itself cannot stop this problem and would only be a first response to protect the population and buy time for more elaborate measures. Not having such measures in place leaves Mexico wide open to macroeconomic shocks.
Robotization and the Coronavirus thus have two further things in common: they can affect anybody, and they are unpredictable. Practically nobody saw the global pandemic coming, and while many estimates are available, how exactly robotization will impact final unemployment figures is insecure, especially for Mexico. The latter is very important when we think about the implementation of Unemployment Insurance, because policies are usually designed with a certain scenario in mind. Normally this is cyclical economic downturns or friction unemployment (seasonal labor, economic bubbles, etc.). If that scenario changes, the policy becomes inadequate. For example, how long should the insurance last? In a country with typical low unemployment such a Mexico, making it long sounds irresponsible in relation to work incentive. But if the labor market changes to a point where there simply are no jobs, longer timelines are needed to allow for retraining efforts. Uncertain times thus require for a flexible design. In what follows I will present you a summary of my conceptual model for Adaptive Unemployment Insurance, written primarily with robots in mind. Take into account this is a summary and all numbers are just examples, not finalities. For a full explanation of all variables and the logic behind each choice, please read the full document available here.
The core concept is to adapt all key variables in the policy when the labor market situation drastically changes. Although they are not exactly the same, we will assume that unemployment serves as a proxy for the impact of structural factors such as robots. This creates four scenario’s (see table):
1. Low unemployment (0-5%): if nothing changes in the economy, we need an insurance that is strict in requirements but sufficient in payment (for example 5000 pesos), since resources are available. The payment is only for a short duration (f.e. 6 months), and the receiver is required to demonstrate a willingness to find and accept work.
2. Strained unemployment (5-10%): What changed, is that more people need help and that they need it for longer. Therefore the amount received should regressively get less each month, for example from 5000 to 2500 pesos. The length of the program is extended (f.e. 12 months) to allow people more time for find new work. The requirements are extended to including retraining, which now becomes a meaningful option.
3. Employment crisis (10-25%): More unemployed with less chances of finding work. Benefits are kept low (f.e. 2500 pesos) to pay for more people, but the time is extended (24 months). Requirements are further broadened to include other useful activities, such as caring for family members.
4. Total escalation (>25% unemployment): society goes into survival mode. The benefits stay low and the time is further expanded (4 years) as reintegration into the labor market is unrealistic for many. The requirements now include volunteer work (harmful for labor market in earlier stages, and of course not in case of epidemic).
The above works for robots and to a lesser extend other economic shocks. For a virus crisis, only the first (two?) levels would apply since it is a short-term event. In this fashion, our Adaptive Unemployment Insurance can both be an interesting expansion of social protection in today’s economy, as a tool for crisis management in change the winds change. A note on funding: these programs are typically paid by a combination of worker contributions, general taxation and taxes on companies (or profits). In a context of robots, placing more emphasis on latter seems ethical, since it are companies who cause this problem and increase profits without creating employment. One could thus increase the percentage of a minor profit (or wealth) tax in each level of escalation, while keeping the other two sources constant. Details such as tax rates are the subject of concrete federal or state level debate and calculation, and don’t concern us here.
On a final note, I want to repeat that although Mexico desperately needs this, it has its limitations. In essence, this policy helps to buy us time, either for the virus to pass or for more effective government responses to unemployment to be unrolled. It also doesn’t cover all people who work, only all people with formal (!) employment. Since Mexican business is allergic to doing things legal, this currently is less than half the population. The government cannot cover those who are not registered, since they don’t contribute nor can prove they lost a job they never (formally) had. If the minds are ready, Mexico could develop a further social assistance program (funded through general resources) that also pays a small sum for a limited duration for all those who ‘claim’, rather than ‘proof’, unemployment. Although an administrative nightmare, such a measure would especially make sense in the current scenario of a virus outbreak. The full proposal, written for the Nuevo León 4.0 network with Robots in mind, can be found on our website.