HOW CRIMINALS RESPOND TO ECONOMIC INCENTIVES: Evidence from the 1995 earthquake in Kobe, Japan

Burglars respond to economic incentives by engaging in fewer crimes when the value of their prospective takings falls. But they don’t seem to compensate for their lost source of income by committing other crimes, probably because of the difficulty of switching to an alternative criminal ‘specialisation’. These are the key findings of new research on the effects of the severe earthquake in the city of Kobe, Japan, in 1995.


The study by Yu Aoki and Theodore Koutmeridis, to be presented at the Royal Economic Society's annual conference at the University of Warwick in April 2019, explores whether criminals respond to a change in economic incentives by analysing how an unexpected devaluation in prospective takings, caused by the Kobe earthquake, influenced the number of crimes committed in Japan.


The Kobe earthquake, which measured a massive 7.3 on the Richter scale, took the lives of over 6,400 people and destroyed over 682,000 buildings. In theory, such a natural disaster could increase property crime by generating a larger need for income. But it could also decrease property crime – and burglaries in particular – by reducing the value of loot as a result of the damage caused by the earthquake.


The authors find evidence for the latter. They argue that, due to the earthquake, the economic returns to burglary decreased, since the damage to buildings and the property within them reduced the value of prospective takings for burglars. At the same time, the prevalence of all other types of crime remained unaffected.


Comparing a set of municipalities hit by the earthquake and nearby municipalities that escaped, the authors find that an increase in the number of damaged houses significantly reduced burglaries. The effect remains strong and significant even after taking account of other factors that may influence the incidence of burglaries, including the number of police officers, the level of unemployment and income per capita.


The study reports the same effect even when looking at the impact of moderately damaged housing on burglaries. This alleviates a potential concern that the earthquake might mechanically decrease the number of potential targets for burglars, and therefore burglaries, simply by reducing the number of houses.


The authors use an innovative method to make a clear identification of the link from cause to effect. First, they show that the larger the distance from the epicentre of the earthquake, the smaller the damage to housing. Then, they compare burglaries before and after in both quake-affected and unaffected municipalities.


They show that the derived effect is stronger closer to the epicentre of the Kobe earthquake. Arguably, the distance from the epicentre influenced burglaries exclusively through the impact it had on housing damage, which confirms the main result of the study.


Other types of crime – such as thefts and other serious offences – were not affected by the earthquake. This probably implies that criminals specialise in certain types of crime, making it hard for them to switch from one type to another.


To illustrate this point using an example from the labour market for legal activities, consider construction workers and compare them with nurses. Even when the job opportunities for construction workers decline, it would be hard for them to switch effortlessly to working as nurses, as the two professions require very different skill sets.


Likewise, even when criminal opportunities for burglaries decline, it would be hard for burglars to switch to being robbers, for example. The criminals would have to give up a set of skills they already possess, such as a skill for unlocking doors, while at the same time acquiring a new set of skills, such as the skill of using weapons.


The authors say that their study provides the first piece of evidence of a response by burglars to the declining value of prospective takings and of criminal specialisation that makes it hard for criminals to substitute out of burglaries into other relatively more attractive types of crime.


They conclude that criminals respond rationally to changing economic incentives in relation to economic returns and specialisation. These important but understudied aspects of criminality shed more light on how illegal behaviour adapts, which can be particularly useful for the design of crime-reduction policies based on rigorous evidence.


HOW CRIMINALS RESPOND TO ECONOMIC INCENTIVES: Evidence from the 1995 earthquake in Kobe, Japan by Yu Aoki and Theodore Koutmeridis


Yu Aoki

Lecturer in Economics at University of Aberdeen

Theodore Koutmeridis

Senior Lecturer in Economics at University of Glasgow