The Labor Market and Information Cascades
An important fact to note is that information cascades do not involve individuals mindlessly imitating the behavior of others. Rather, they occur because individuals draw rational inferences from limited information, a major part of which includes observations about the behavior of others. This phenomenon also applies to the labor market where employers receive signals about the abilities of employees by looking at their employment histories.
For instance, a good job in the past would imply that previous employers were happy with an employee’s abilities, sending a positive signal to employers in the future and making it more likely for the candidate to get hired again. On the other hand, a bad job in the past would imply that previous employers were dissatisfied with an employee’s performance, thereby sending a negative signal to employers in the future and hindering an employee’s chances of getting hired. The same applies for periods of unemployment that send negative signals to employers about the abilities of a worker.
Now it is important to see how signals transmitted about the qualities of workers in this manner cause information cascades, of both positive and negative natures for employees. Positive cascades can be referred to as “up cascades” whereas negative cascades can be referred to as “down cascades”. “Up cascades” occur when one good job for a worker transmits positive signals to a subsequent employer who hires the worker directly based on the signal. On the other hand, a “down cascade” occurs when a spell of bad employment or unemployment transmits a negative signal to subsequent employers who do not choose to hire the worker based on the signal. Thus an “up cascade” can cause a worker who receives good job offers at the beginning of his/her career to become very successful in the long run. Similarly a “down cascade” can cause a worker with a bad career start to remain unsuccessful in finding employment with subsequent employers.
It is interesting to note, however, that “down cascades” are a lot more indicative of a worker’s true abilities than “up cascades”. This arises from the fact that workers are able to manipulate signals transmitted between employers. Take a bad worker type, for instance. Given that a good career start could lead to a successful long-term career, such a worker has a high incentive to initially exert higher effort than normal to trigger an “up cascade” in his career, in order to manipulate signals transmitted to employers (for the better). This makes signals generated during “up cascades” less reliable, because they are not always reflective of the true nature of workers’ abilities. On the other hand, a bad worker type does not have as high an incentive to avoid a “down cascade” as such a worker is less likely to succeed in the future anyways. At the same time, a good worker type would generally not be subjected to “down cascades” because of his/her inherent abilities and hence, the quality of signals transmitted during “down cascades” would be of a more reliable nature in both cases.
Therefore, the aruments above reinforce the transparency of “down cascades” and make them more reflective of real information, as opposed to “up cascades” that have the potential of getting influenced by the manipulation of signals transmitted.
– AS2411
Source: http://www.wm.tu-berlin.de/fileadmin/fg210/Publikationen/infcasc.pdf