Reply to Ariel Rubinstein's critique of “A fine is a price”

Recently (Saturday, August 20 th 2005) Ariel Rubinstein was a discussant in a session on Behavioral Economics in the World Congress of the Econometric Society. In his discussion, Rubinstein criticized, among others, our paper “A Fine is a Price” ( Fine.pdf ). Since we were not there we could not reply to him directly. We now respond here instead.

Rubinstein raised two objections. We list them below followed by our responses:

Israeli managers would not collect the fines : The practice of giving an NIS 10 fine is part of the options that managers in those daycares have. They all belong to a private kindergarten organization, which recommends to the managers either to give no fine or a NIS 10 fine. As an empirical fact, the managers in the daycares who participated in our study collected the fine.

The way the data was collected is inappropriate : As a common practice in those daycare centers, assistants and managers wrote down who came late every day. On this basis the managers charged the family the total fine at the end of the month as part of the day care center bill. Our RA collected this information every Friday. These are the data we report. We wish to emphasize that the managers and the RA were not aware of our research hypothesis. The procedure, as well as well as all the data collected, is reported in details in the published paper.

This is not the first time that Rubinstein raises these objections. He did so also six years ago, before our paper was published. Our answers did not satisfy him, and he advised us to withdraw the paper from the Journal of Legal Studies before the publication. Since we were fully convinced that the design, the method, the data collection and the result were sound, and that the results were therefore reliable, we eventually declined to follow his advice.

We believe that the ultimate test of the concerns raised by Rubinstein should be a replication. We offered Rubinstein to take part in the design of such follow up study that included a replication of the original one. Rubinstein declined. Nevertheless, we did run the follow up study six years ago. There is a second day care study—that fully replicated the results of the first study. The results of the replication are at 2DC.pdf .

In the follow up study we asked the managers to allow us to give Rubinstein their phone numbers so he could confirm with them that the design was followed.

Rubinstein's does not present factual evidence that supports his criticism of our experiment. We did check the objections and found that the managers in our study did collect the fines and that the way in which we gathered the data was accurate.

Uri Gneezy and Aldo Rustichini .