Dr. Simeon Djankov:

It sounds very depressing. That does not mean we should not try. Chapter 4 of the report goes into detail on the reason there is such a negative return for such investment. Some of it we can rapidly change. Many adult retraining programmes - I am familiar with some of the European programmes as they are applied in my country of Bulgaria - involve bringing in adults and treating them like students, seated at a school desk and giving them courses spanning three, six or nine months. They take classes as if they are students. We spoke about young brains being able to learn technology much better but adult or more mature brains operate very differently. This is the area of neuroscience. In particular, on average, every 1.7 seconds an adult brain switches to some other gear. In other words, for adults, their unit of attention is less than two seconds. For children, it is more like 20 seconds. We need to organise the course material or teaching methods for adults very differently from the way we organise them for children. Adults need much shorter learning units, not 45 minutes but 20 minutes and the use of much more technology. A number of programmes we use that have positive returns literally comprise 15 to 20 minute sessions over the course of an hour and a half or a two-hour training day and texts are sent to the participants during the latter part of the day asking if they remembered a matter and that is covered in just one sentence. That model works much better. Lithuania is one such country that uses European Union funds to retrain in that way. With approximately one and a half hour courses a day over a two-month period, its rates of return are significantly higher than a ten-month course in Germany, provided for exactly the same purpose, for example, to retrain low-level engineering skilled workers. We can improve adult retraining programmes simply by not treating adults as older students. Beyond that, the analysis suggests that by the time one is 30 or 40 years old, one's ability to switch to a new sector is fairly limited and, therefore, one tends to switch to a neighbouring sector. It is not possible, for example, for a truck driver to become a software engineer or a marketing agent. We need to think step-wise how workers can move from one job, which is under threat of automation, to a closely-related job that has a longer-time horizon. A few countries have succeeded in that but there are very few. The Senator is right in that the evidence is not yet optimistic.