Like many Venezuelans, Camila checks her social media accounts twice daily to know the exchange value between the American dollar and her country’s local currency, the bolivar. Since 2014, Venezuela has experienced an unprecedented social, economic, and political crisis with the highest levels of inflation in the world. At its peak in late 2018, the inflation rate was 130,060%, with the percentage staying in the hundreds ever since.
Camila must check the American dollar’s value several times per day because the Venezuelan bolivar depreciates so quickly that its value differs from morning to afternoon. She earns money by working from home, labeling data for machine learning companies that develop self-driving cars, content moderation algorithms, and automated voice assistants. In a country where 51.7% of the workforce are freelancers, this remote work allows many like Camila to earn an income in American dollars and mitigate the hit of hyperinflation on workers’ personal and household economies. However, platform-enabled work at a time of economic crisis and uncertainty also creates dependency and precarity.
In a country where 51.7% of the workforce are freelancers, this remote work allows many like Camila to earn an income in American dollars and mitigate the hit of hyperinflation on workers’ personal and household economies.
Like other data workers, she generates machine learning (ML) data. As an example, one client asked her to take photos of herself and to send photos of her childhood to train an algorithm that predicts aging. She also annotates data, giving meaning to the binary ones and zeros according to pre-defined classifications. For a self-driving car, she would label the objects in images of roads, distinguishing humans, buildings, and vehicles, among other categories. Finally, she would also verify the accuracy of algorithms and evaluate their predictions, notably to improve search engines. These variable tasks constitute the core of data work outsourced through digital platforms that have targeted Venezuela because of its cheap labor and digital infrastructure.
Many platform companies started to recruit workers when the crisis began in 2014. The ecosystem of platforms constantly fluctuates because of mergers and acquisitions and even bankruptcy. Earlier platforms like CrowdFlower and Figure Eight gave way to more recent ones like Appen, Hive, and Microworkers. More recently, the Canadian company Telus has become a main player in this market. These platforms provide services to numerous technology companies from larger corporations like Google and Meta to smaller start-ups. Additionally, at a time when more sectors of the economy are incorporating ML into their products, these platforms’ clients include a variety of companies such as car manufacturers, logistics companies, and appliance makers.
Camila learned about this type of work from acquaintances who spoke of the possibility of earning an income in American dollars from the comfort of home, only requiring a computer and an internet connection. The platform pays her USD 10–20 per week, an income much higher than the local minimum wage, which was around USD 1 per month in 2020. However, this income is barely sufficient for her and her family to afford necessities every week, especially after considering all the money she must pay off in fees and transactions to all the financial platforms she uses.
Financial transactions in digital work are complex since several intermediaries are present before workers can receive their wages. Labor platforms pay workers through platforms like PayPal and the Mexican app AirTM. Each of these services has its rules and operations. They require workers to authenticate their accounts by uploading their identification cards and matching the photo in the document with a selfie using a third-party facial recognition service. Transactions through AirTM are in a digital currency pegged to the American dollar called the AirUSD, and, as a peer-to-peer financial platform, the money stays online. If a user wants to cash out their account balance, they must sell it to another platform user who pays by transferring the cost through other channels.
The nature and restrictions of PayPal and AirTM require workers to use another layer of financial services to receive their wages. Many workers use other electronic wallets (or e-wallets) like Binance, Uphold, and Zelle to transfer digital funds, and especially, for transactions with cryptocurrency. Some data annotation platforms have begun to offer crypto like Bitcoin as a form of payment, compelling many workers to learn more about these financial assets.
Furthermore, for some workers, crypto has also become an investment opportunity, notably before the pandemic, when the market value of Bitcoin and similar currencies constantly rose. At a time when hyperinflation has drastically diminished investment opportunities locally, many online workers see Bitcoin and other cryptocurrencies as an alternative with high returns and easy access.
Venezuelans who work online learn about these financial platforms and bear the additional labor of transferring their money across them because of their low access costs, the possibility of earning their income in American dollars (a more stable currency than the local bolivar) and being able to save and invest without the value of their assets diminishing rapidly because of hyperinflation. That said, the costs make this arrangement far from optimal.
Each one of the services used — from the instant the labor platform pays the worker to the moment they cash out their account balance — profits from transaction costs. In addition, workers must spend additional time learning about, setting up, and using these financial services. As independent contractors or freelancers paid per piece, the financial aspects of their work take a considerable amount of time away from their productivity, as well as money from their earnings.
Each one of the services used — from the instant the labor platform pays the worker to the moment they cash out their account balance — profits from transaction costs.
Previous research on online platform labor has found that gig workers had to bear additional risks and costs associated with their activity since the Covid-19 pandemic started. In addition to health and social risks, this distribution of wages across platforms creates financial risks for workers. The recent collapse of the value of many cryptocurrencies, where many workers have invested and saved, shows that the promise of high returns of these distributed digital assets also bears high risks, especially for low-income investors worldwide like Camila and other Venezuelan data workers.
Platforms also fall short of providing a complete solution for online wage payments. Many workers still require local currency buyers or brokers to convert their digital funds to bolivars. While workers learn from and contact many of these brokers online, they are often recommended by trusted close connections such as family members, neighbors, and other Venezuelan data workers because of the high risk of scams and online criminality regarding digital transactions.
The risk of fraud is so high that the most effective form of worker organizing that has emerged in Venezuela around online labor platforms revolves around financial transactions. In my research, I found several worker communities organized on messaging platforms like Telegram and WhatsApp. These groups are formal, with membership fees and application procedures that involve disclosing the applicant’s identity, including identification cards and home addresses.
The case of wage payments in Venezuelan data work shows that nothing is entirely ‘virtual’ and that the flexibility and ease of the online world comes at a cost. Indeed, the arrival of online labor platforms has provided an opportunity for many Venezuelans like Camila to alleviate the social and economic problems associated with their country’s ongoing crisis. Thanks to the many digital platforms — for labor, financial transactions, messaging, and other applications — Venezuelans can organize, work, and earn an income in a more stable currency, as well as save, and even invest.
However, regarding financial transactions and the platformization of their wages, the more workers become embedded in these networks of firms and other local social actors like brokers, the more costs they must bear. Every node in that network provides a service, each of which has a price that diminishes returns for workers and depreciates their wages. They prove that flexibility and easy access have costs and risks.
The recent collapse of FTX and the depreciation of Bitcoin also show, that more than currencies, crypto is a digital speculative asset that has, for the most part, benefited those at the top of the scheme to the detriment of the more precarious users, like freelancers. These risks are higher for those living in a social, economic, and political crisis, who do not have any other option but to depend financially on electronic wallets, online employees, and all the actors involved in deeply embedded digital finance.
The recent collapse of FTX and the depreciation of Bitcoin also show, that more than currencies, crypto is a digital speculative asset that has, for the most part, benefited those at the top of the scheme to the detriment of the more precarious users, like freelancers.
Ultimately, digital payments are far from a techno-libertarian dream of transactions without institutional oversight. Like in many other instances of platformization, in which the transnational and digital nature of the companies’ intermediation translates into poor regulation, the firms behind the digital infrastructures reap most of the benefits of the enormous power differentials. E-wallets get to set their rules and fees only subject to the market, and workers must defend themselves from losses and scams.
Despite the benefits, there is a cost in terms of dependency and loss of autonomy associated with these digital platforms, from financial to labor intermediaries. Like a double-edged sword, they provide opportunities but also risks; a way to circumvent a crisis but also a new mode of exploitation of a precarious workforce. These platforms are tools to survive a crisis perceived as temporary, but they also allow access to dependent low-income users who are only considered as accounts, ones and zeros, whose labor is but a commodity. The question lingers as to whether this arrangement is here to stay in the long term and how to make it more sustainable in the future.