Situation: Between 20% and 30% of the working-age population in developed countries are freelancers. The Financial Times predicts that by 2050 this figure could be as high as 83%. The vast majority of those who choose to work freelance use digital platforms to find jobs. The online labour market gives freelancers the freedom to manage their time, they cannot be 'locked into' working hours, and the privilege of the self-employed is to determine their own hours.
In fact: In reality, there is the so-called ‘autonomy paradox’ or 'myth of flexibility'. To get work in a highly competitive environment, freelancers adapt their own routines to the needs of their clients, so they have to work long hours not only during the day but also during non-standard hours, obeying the unwritten laws of online platforms.
Denis Strebkov, Andrey Shevchuk and Alexey Tyulyopo from the Laboratory for Studies in Economic Sociology (LSES) of HSE University investigated how online platforms shape the work schedules of freelancers and what consequences this has for them. In addition, they clarified the motivation of freelancers themselves — what drives them to work evenings and nights, and whether everyone (regardless of experience and qualifications) switches to non-standard hours in the same way, as well as the role played here by the boundaries created by time zones. For this analysis, scientists used data from the largest Russian-speaking online remote work exchange platform, FL.ru — 241,600 messages posted by 29,800 unique users in 4,082 contests. The results of the study are published in the journal New Technology, Work and Employment.
A significant part of freelancers' employment during non-standard hours is ‘invisible’, i.e. unpaid work that only allows them to take orders. This includes, among other things, networking.
You have to act promptly, responding to requests at any time, and for some it is important literally not to sleep away their earning potential. We are talking about those who are in different time zones to their ‘employer’, and if the difference between them is significant, the chances of a resident of the Far East, for example, losing out to competitors from Central Russia, increases.
Freelancers who are not 'local' to the client are at a particular disadvantage. Such cooperation requires them to adjust their schedules and shift their working hours to non-standard times. Both the direction and the number of time zones crossed matter: those living to the east of the business centre have to work in the evening and at night, while those in the west have to work early in the morning.
The situation is also characteristic of the global online market. The majority of customers are from the US, Canada and Western Europe, and the majority of employees are from Eastern Europe and Asia. For example, call centres in India operate at night to serve customers from the Global North – developed countries that are 12 hours behind.
Russian-speaking freelancers are a separate object of study. FL.ru has 1.5 million registered users from 20 time zones, including Moscow (GMT+3). This coverage has helped researchers to break down the problem in detail, using quantitative data, which is still lacking on this topic in Russia and the world.
The authors considered three time slots as non-standard:
non-working hours — from 6 p.m. to 9 a.m.;
night hours — from 9 p.m. to 6 a.m.;
weekends — Saturday and Sunday.
The extent of the work was measured by analysing the proportion of messages sent by freelancers during these periods on their local time, out of their total number of messages. This is the 'invisible', i.e. unpaid activity in which job seekers submit tenders for proposals and communicate with clients on work-related issues.
The main independent variables are the proportion of customer messages posted at non-standard times from the freelancer’s point of view within each tender and the time zone in which they reside.
Variables that were controlled for (qualitative characteristics capable of influencing the measured values) were gender, age, place of residence, qualifications and reputation of freelancers, as well as information about 4,082 contests conducted on FL.ru by 3,063 clients in four years — from October 25, 2014 to October 6, 2018 (remuneration, duration, number of participants, type of task, etc.).
Workers’ characteristics were taken from profiles on the platform, and employment schedules were tracked by ‘digital footprints’, i.e. information about users (their behaviour and online communication) recorded by the website.
A tender on FL.ru (usually for the creation of a brand book, logo, package design, slogan or company name) attracts an average of 45 entrants, lasts 15 days and promises the winner around $200-$250.
Clients in the study: 91% are from Russia, including 52.7% from Moscow and 12.1% from St. Petersburg.
Freelancers: from 1,143 localities in 82 countries. The majority (76.6%) are from Russia, 11.5% live in Ukraine and 9.4% live in other post-Soviet countries.
The study confirmed that the transition of freelancers to non-standard schedules depended on client activity. The more the clients posted on weekends, in the evening or at night in terms of the freelancer's local time, the more the freelancer worked during these particular periods.
Presence on the platform between 6 p.m. and 9 a.m. takes up an average of 42% of a freelancer's time. If clients are active during this period (their messages are posted), the freelancer’s activity increases by 11 percentage points to 53%. On Saturdays and Sundays, the effect is even stronger: +31 percentage points, from 15% to 46%.
‘Freelancers are very responsive and receptive to customer behaviour,’ the researchers say, ‘On average, they respond to messages five times faster than they receive replies. This suggests that freelancers are constantly on 'standby' and are the ones adapting to customer requirements, not the other way around.’
Almost 2% of the contests received an initial response in less than a minute and more than 50% within 10 minutes.
Overall, excluding the initial response (tender application), the average response time of a freelancer in correspondence with a client is 1.8 hours. The client response time is 9.4 hours. During out-of-hours and weekends the response times are 2.7 and 13.4 hours respectively.
When tenders are posted and every minute counts, the time zone factor comes into play. If it is published at the end of a Moscow working day, those who are in the capital city time zone are at an advantage. The online platform in this sense does not level the playing field, as is commonly thought, but rather discriminates against people from the time ‘periphery’.
Each time zone east of Moscow increases a freelancer's share of non-standard and night work time by 1.5 percentage points, in other words, forcing him or her to go to bed 14 minutes later. It turns out that a freelancer from Western Siberia ends the working day one hour later on average than a Muscovite, while a freelancer from the Far East ends the working day two hours later.
Men, young people, and Russians have the highest propensity to work irregular hours. Experienced freelancers have the least. The more experience and higher qualifications they have, as well as their ranking on the platform, the weaker their response to customer activity, including job postings.
The old-timers, compared to the newcomers, ‘work into the night’ only out of necessity, mostly responding to client letters, while the newcomers spend a lot of their time searching for and submitting applications.
The freelance economy (gig economy) is gaining a firmer foothold in the employment space, but freelance work is not only convenient but also controversial and problematic. To find out what the problems are means understanding how the online labour market functions and affects people. HSE University’s study is an important step towards understanding the topic, one of the few based on a large-scale sample, and the first to focus on the ‘invisible’ work of freelancers.
The mechanisms of digital platform pressure on a self-employed person's work schedule are also not a frequent subject of scientific interest. However, studying them can prove or disprove the ‘flexibility myth’. And analysing work in an irregular schedule is able to help answer questions about maintaining freelancers' health, well-being and family life, since it has been proven that superhuman efforts on the internet do not add to health or personal happiness.
The dataset used by sociologists at HSE University is in the Russian medium, which is also extremely important. Online labour exchanges in Russian reach an audience from Russia, post-Soviet territories and Russian-speaking people around the world. This is a huge market through which the nuances of the gig economy can be discerned. At the same time, it is largely ignored by researchers: most monitoring today is concentrated on English-language platforms with customers from rich countries and workers from poor ones.