Do women prefer longer conversations? Evidence of communication preferences from commercial and private telephone use

We investigate whether there are systematic differences in communication behavior between men and women of a kind that could affect their propensity to form strong versus weak social ties. We also investigate whether such differences result from different preferences rather than from different constraints or simply from different habits. We report two sources of evidence on telephone usage, to test the hypothesis that women prefer to communicate with fewer partners but prefer to spend more time with each. First, we report a study of anonymized billing records of a random sample of 3103 subscribers to a large mobile operator in Italy and Greece over two years from 2006 to 2008. This shows that, when faced with identical tariffs, men and women have different calling patterns, with women making fewer but longer calls than men on average. Calls made by women are around 16% longer on average when other factors are controlled for. The findings are highly statistically significant and are found across all age groups in both countries. Secondly, we report a study of some 92,000 person-days of calls to call center employees of a large consumer services company operator at four sites in Germany. Calls received by women employees last around 15% longer than those of men when other factors are controlled for. Again the findings are highly statistically significant and are found across all age groups. The random allocation of incoming calls to employees makes it extremely unlikely that this is due to differences in calling incentives or opportunities. We also show that these gender differences are weaker when communication costs are high, which suggests that they represent different preferences for calling behavior rather than ingrained habits. Our data are silent on whether such differences are innate or are due to early socialization of boys and girls. We discuss possible implications for other aspects of gender differences in social life, including clustering by gender in the labor market. 1 The authors are grateful to Nicoletta Berardi, Michael Kosfeld, Jonathan Sandbach, Neele Siemer and Levent Yilmaz for comments, help and advice. The usual disclaimer applies.