A new study sought to shed light on the question by examining the records of hundreds of workers at two clothing factories to see if the makeup of a workforce had any impact on the productivity of the factory.

Using data from the equivalent of almost 35,000 worker days, a group of academics found evidence that even small increases in the number of social contacts in the workforce can have a significant effect on individual and group productivity.

The study, “The ties that bind us: Social networks and productivity in the factory,” was published in the Journal of Economic Behaviour and Organisation and co-authored by Professor Amrita Dhillon, from King’s College London; Professor Farzana Afridi and Swati Sharma, from the Indian Statistical Institute.

Data for the study was gathered from two garment factories in the state of Haryana, India, over a period of 31 continuous working days, with sample data from 1,744 workers and the equivalent of 34,641 worker-days.

The academics surveyed workers to gather information on individual characteristics, including current residence, native state of residence and caste, years of experience in the garment industry, and the process of obtaining the current job, particularly referrals.

Data was then collected on worker performance and worker attendance over a period of 31 consecutive working days. Workers were typically employed on production lines and were responsible for stitching garments.

Because of a limited supply of skilled labour and the high proportion of migrant labour in the industry, worker absenteeism and turnover was high. Academics found a 31 per cent average daily deviation in the number of available workers, meaning they could observe and measure the impact of changes in the makeup of the workers on productivity on a day-to-day basis.

The academics found that a one percentage point increase in the proportion of workers of one’s own caste saw a boost to the average individual worker’s efficiency by at least 0.09 percentage points, with the least efficient worker’s productivity rising by about 0.17 percentage points.

The academics noted: “Our findings demonstrate the importance of pre-existing social connections in the form of caste-based networks, amongst workers as another channel through which economically interdependent workers can influence each other's performance and thereby affect the group output.

“Our results indicate that deep social interactions outside the workplace carry implications for productivity - with corresponding implications for optimal design of production schedules and composition of teams - within the firm.”