Theoretical Speculations
In Bologna many people do not buy the new bikes, instead they tend to buy the old and stolen ones and almost no one cares whether it is stolen or not, i.e., it is not considered morally bad to buy a stolen bike at least among the students. It seems that the reason is the prominence of the bike theft in Bologna. In such a context with a high probability of one's bike being stolen, in making a decision for buying a bike, as the presumption of rationality suggests, nobody should buy a new one. Two other strategies left are either to buy an old non-stolen bike which is supposedly more expensive or to buy a stolen old bike. Since no one can track a stolen bike (there is no numbering or identity card (registry) or any other electronic tracking device for them) and also the buyer runs the risk of getting his bike stolen, s/he will prefer the stolen bike which is supposedly cheaper (at a discount, let's call it the theft discount).
This seems to be working very well and everyone is happy, though those who attach some sensational values to their bike certainly are not. Why is it so? First: It does not cost a small fortune for the buyer to buy the bike and since the buyer did not pay much for the bike, s/he is not regretting the bike getting stolen. The thieves are taking the role of the brokers, and they are bearing the risk of being caught by the police and hence are getting compensated for bearing the risk with getting the whole price of the stolen bike. So the thief is happy too, because he is getting a brokerage fee and living a life on that. In this way, it seems that in the absence of state enforcement of property rights or let's say its weak enforcement, people are getting around and reaching (for now) an efficient outcome. Thought deciding whether the situation is efficient or not requires more investigation.
The problem with this analysis is that the original buyers are the real losers and no one will buy a new bike. So, if we can generalize this non-enforcement of property laws, the bike making factories will go bankrupt. But it seems there is a counter argument to this too. Since the bikes will be worn out with the passage of time, and they will not be usable anymore, there will be new demand for the new bikes, even at the higher prices. But still there is hold-out problem, not anyone will be the first mover to buy the new one because of a huge first-mover disadvantage, i.e., because if he buys it at a higher price, he will run the risk of losing a fortune. This seems to be a dead-end to the this speculation, but still there are people who value the new bikes more and they might prefer having a new bike despite knowing that it might be stolen. In addition, the hold-out problem will contribute to decreasing the price of the new bikes down (it will decrease the demand for the new bikes). There is another concern with that and it is a very well-known problem called "Race to the bottom". Since the bike factories will come across the shortage in demand, they will produce low quality bikes so they can survive. Therefore, one of the biggest problems is the decreasing quality of the bikes produced by the bike factories.
Besides, it might also be said that the buyers of the stolen bikes are free-riders, because they do not have to pay the full cost of bike which is paid by the original buyer, but given that the buyer of a stolen bike runs the risk of his bike being stolen, s/he should not be taken as the free rider, s/he just gets the bike at a discount for the risk of being stolen., or to put it in another words, s/he probabilistically pays for that.
I am also told that if your bike is stolen, the victim can go to the police and by showing his or her bike to the police, request the removal of the lock and the return of the bike to the original owner, the police will break the lock and give the bike to the owner. This, which I am not sure of its credibility, sounds very stupid; the problem with this solution is that when one goes to the police, how can the police verify his statement about the ownership of the bikes? Without traceability and registry system, it is quite stupid to go for such a solution.
One might say that there will be asymmetric information and adverse selection problem between the thieves and the buyers of stolen bikes. Certainly there will be such a problem and this will cause inefficiencies in the market for stolen bikes. This asymmetric information will create a situation in which the thieves should either issue an insurance policy for the stolen bikes or issue a warranty for them while selling them or sell them with a discount. Since there will be no enforcement for such an insurance, no one will buy that insurance. Hence there remains the possibility of discount and in fact, the thieves are actually selling the bikes with a huge discount. This huge discount causes another acute problem.
Say the police wants to increase the bike theft enforcement to a full enforcement so that there will be no theft or next to zero theft. In this case, this level of enforcement will affect the discount rate in selling the stolen bikes, that is, this high level of enforcement will increase the discount rate. The increase in the discount rate in turn will affect the enforcement of the property rights with regard to the bikes, how? With increasing discount rate, the prices of the bikes will reach to a very low level that the very enforcement will be inefficient. It means that this level of discount will wipe out the efficiency of the enforcement. Though this enforcement level might reach to a point in which the very theft might be inefficient too. By a strict enforcement, the police can eliminate the market for stolen bikes, and the thieves seeing this will not steal the bikes at all, but it might be inefficient to do so.
The question is that if we can model this phenomenon as a game or design an experiment in a game theory setting for this phenomenon, which set of payoffs, will be the Nash equilibrium? Is this move from an enforcement based property rights to a so to speak market based property rights free of enforcement an optimal Pareto-efficient outcome? Is there any concern with regard to the justice and fairness in this example? What other consequences this model might have?
The Experiment Design
There are two hypotheses in response to the question of “What causes the bike to be stolen more in a city like Bologna and less in other cities?”:
1. First hypothesis is the law enforcement is less with regard to the property rights in bikes in Bologna than other places.
2. Second hypothesis is that there exists a positive relationship between drug addiction and the number of stolen bikes.
So, beside those theoretical concerns about the bike theft, what I am going to do in this short draft of a proposal are two things. First we will see if there is any relationship between non-enforcement of property rights about the bikes as the independent variable and the rate of the stolen bikes as dependent variable and second since there are some speculations about the drug addiction and its rate on the rate of bike theft, we shall by experiment see if there is a relationship between drug addiction as an independent variable and the stolen bikes rate as dependent one. This could be done either within the experiment while selecting the subjects heterogeneously, i.e., the drug addicts and the drug-clear subjects or by choosing a completely homogeneous group of drug addicts and designing a new experiment for that purpose.
To design the experiment and in order to be able to measure the differences that might arise from the different experiments done, it seems we should divide our sample groups into two groups, i.e., the experimental group and the control group. Control group is the group which is not exposed to the independent variables, instead, the experimental group is the group who is under the experimental variables and we expose them to the independent variables. So first we should develop an experiment with zero level of enforcement and then by changing the level of property rights enforcement (civil and criminal sanctions) measure the response of the behavior of the thieves to the level of change in the stolen bikes.
To do that we shall first see what are the constituent parts of the behavior in this setting and what affects the behavior of the subjects. Intuitively, it seems that the followings are contributing to the behavior of the subjects.
1. The level of enforcement which can be manipulated by, for example, a. installing the surveillance cameras in the areas in which the bike racks are located. b. making bikes traceable through numbering or special registration or c. making people sensitive about it which will have the force of a social enforcement and peer pressure or to put it differently, making the social enforcement more stringent through which making the secondary market for the stolen bikes illiquid. This illiquidity problem may have an extraordinary effect in the incentives of the individual thieves. This can be done, by issuing a certificate for every bike, in the experiment and then banning the people from getting the bikes without the certificate. After each step, the re/actions of the subject should be subtly recorded.
2. The price gap between new bikes and stolen bikes. This price gap, which is huge in reality, can incentivize the thieves to get more involved in the theft of bikes. In our experiment, by filling this gap or widening the gap, one can manipulate the incentives of the thieves to steal a bike or not.
3. There might be a host of other variables contributing to the theft behavior which can be measured while selecting the individuals as the experimental or control group. Though not exhaustive, the characteristics which might contribute to the theft behavior are as follows: economic status (occupation and earning power), drug addiction, parents’ occupation, the size of family, religious affiliations, schooling and education, and a host of other characterizes.
To do the experiment, the following procedures should be followed:
There should be a number of subjects representing the social mosaic of the society in question, to do so; we shall take two steps, first steps is to decide to select the subjects either randomly, or from a pool of people knowing their social standing. In this stage, it seems randomization can work better, it should be stated that in the process of recruiting the subjects the public announcement should not be used, because it might generate biases, like self-selection bias. That is mostly because if we use public announcement for recruiting which is supposedly paid for the experiment, it will urge relatively poor people enter into the experiment. In order to check to see if there exists any relationship between certain characteristics of the subjects with the level of the bike theft, we may use a questionnaire to check their backgrounds and have access to their background information. Though this may blur finding the causality relationship in our experiment, the best way is to select subjects randomly from a melting pot and preferably from different backgrounds and meticulously track their behavior the using a questionnaire check for their social standing and overall background.
Since one of the important elements in criminology is the probability of being caught and punished by the police/law enforcer, i.e., higher or lower level of being subject to the law enforcement, it will not be enough to make laws or rules in this context and publicly declare to all the subjects that the punishment for this crime is high, if the level of the punishment is high, but the level of the enforcement is low, the thieves who are rational decision makers, will calculate the expected utility of the crime, that is they will calculate the disutility of being caught by the police., in monetary or non-monetary terms and will compare it to the utility they will derive from stealing a bike by multiplying it to the disutility from the probability of being caught.
Repetition matters: repetition will matter, if the design of the experiment is game theoretic. As two examples, reputational effects and the credibility of threats can be included in this setting. If the police establishes a reputation of being tough, this will affect the behavior of the subjects in the next repetition. This can be said about the credibility of the threats imposed by the police to take tough action in dealing with the thieves. In this setting, the peer pressure will become increasingly important with the repetition of the game too. That is to say, by creating a belief that the theft is morally bad, or internalizing the belief in all subjects that the breaking the law is bad, there will be a peer pressure on the subjects who want to steal the bikes. While stealing there would be a sense of being demoralized and also fear of voluntarily report of the theft by the third parties/peers. By repetition, anonymity and non-anonymity of the subjects will matter, taking into account the fact that the theft might be considered immoral; we should design a setting to eliminate the effects of the peer pressure or anonymity to successfully measure the relationship of the law enforcement with the rate of theft.
The condition of the bikes might matter. The newer the bike, the market for that will be more liquid because the buyers can take more risk to get the bike even with higher level of law enforcement. In addition, the price of the accessories to take care of the bikes, such as locks might matter as well. Sometimes the price of locks surpasses the price of the bikes themselves, so bringing the utility of the older bikes lower to be bought in the market for stolen bikes.
To find the relationship between the drug addiction and the level of bike theft, changing the price of the drug as an independent variable could be very helpful. Since there are speculations that the price the thieves charge for a stolen bike is a function of the amount of the drug good for one time consumption, changing the price of the stolen bikes in the market might have a substantial effect on the incentives of the thieves who are drug addict. The questions that should be answered in the experiment with regard to the relationship of drug addiction and the bike theft are as follows: Will the change in the price of the drug affect the price of the stolen bikes? If the price of the drugs is much higher than the price of the bike, will it worth for the thief to steal the bike for buying some drug? This might be tested by changing the level of the price of the bikes too.
