Shyam Krishna| PhD Candidate| School of Business and Management at Royal Holloway, University of London.
In recent months amidst the uncertainty of COVID-19, food delivery workers in India have faced a tightening of their working conditions. The main contention voiced by the workers in strikes and protests is that digital platforms have lowered effective wages by changing the payment structure and its underlying algorithmic calculations. My research seeks to understand the uncertainties in how wages and other work conditions are affected by algorithms. As part of a research project (Investigators: Dr. Yingqin Zheng / Shyam Krishna ) I queried such algorithmic ‘social justice’ implications of digital platforms. Set in the south Indian city of Chennai, I conducted several interviews with food delivery workers (called ‘riders’) and engaged in the work myself in a form of auto-ethnography. I created as many opportunities as possible to observe first-hand the algorithmic control of work, the customers, the digital platforms, and its processes, and also the restaurants using these platforms. I worked on two different food-delivery platforms. I found that a steep learning curve faced me which entailed learning to navigate the digital platform, gaining local knowledge of the city and even knowing the local cuisine.
Working as a rider over a period of five weeks just as the COVID19 crises was blooming globally and in continuing engagement with other riders during the pandemic I gained some useful insights into the ‘gig-work’ practices with some being specific to the Indian context. Mirroring recent work on ‘spatiotemporalities’, the main contestation as found in my research is between workers and the platform in how their ‘space’ and ‘time’ are algorithmically controlled and manipulated.
An aspect of spatiotemporal negotiation emanated from the first order I delivered. The order was assigned to me with a rather resounding buzz on my phone
All platforms use haptic feedback and a loud alarm on the smartphones designed to grab attention and ultimately control rider behaviour to attend to orders quickly. I had about 30 to 60 seconds to accept the orders assigned to me without clear information of the distance to be driven or the address for delivery. The only information I saw on screen was the estimated time taken to reach the restaurant. Features such as alarms and partial information shown are ostensibly designed to add pressure and even panic at many points during the food delivery work process. These temporal pressures play out during riding on the road which itself posed significant challenges specific to an Indian urban context. The risks on road due to traffic were compounded by near constant exposure to air pollution or difficult weather conditions during the hot and humid days in Chennai. Such risks are transferred from the customer to the rider as is inherent to the gig-work practice and become a common expectation of work conditions that workers navigate.
I barely juggled such vagaries of working alongside the technical issues I faced – such as using mapping services as location-based-services were often imprecise. Extra but unpaid effort was required to sort out the erroneous approximations automatically generated by the digital platform or manual errors made by customers in marking locations on a map. Moreover, it was repeatedly reinforced in training that the closest rider to a restaurant (in theory at least) is assigned the next order. So, there was a quite a lot of effort in figuring out the correct and the most optimised location. Luckily – or as is probably the way in which many new riders learn this – I was helped along by other experienced riders. Some of these riders even took the time to escort me to specific places and gave me tips on the time of the day to arrive there.
Moreover, algorithmically defined but imperfect estimates of waiting times or delivery times were a constant issue faced by workers and are referenced by the restaurants and the customers even when we met face to face. Power and information are privileged to customer and restaurants, the platforms use the riders in their subordinated position to negotiate difficult physical conditions arising within the digitised food delivery process. This happens under the close control and manipulation of workers’ space and time even when such extra efforts are unpaid and unaccounted for in how wages are determined.
The findings from this research suggest that riders face unfair conditions and intensive control of work broadly brought about by time-controlled and location-driven algorithmic elements. There are clearly intimate and individual spatiotemporal machinations within gig-work under the mostly opaque nature of platforms and algorithms. These have been conceptualised as ‘spatiotemporal (in)justice’ to probe the aspects of (un)fairness and (in)equity faced by gig-workers. Centring on spatiotemporal justice then would help establish what ‘fair’ pay and practices, standards, and metrics for food delivery gig-workers might look like – which forms the basis of the ongoing collective efforts within India and beyond. The research project itself has resulted in a report on unfair practices in food-delivery work shared with riders, labour leaders and community organisations in Chennai, to assist in ongoing efforts. A further academic paper is forthcoming in the IFIPJWC 2020 conference proceedings.