Episode 12 | Entrapped by E-Commerce

Life before e-commerce can feel like a foreign country. For the ‘Chinamaxxing’ crowd, fresh meals and fancy gadgets delivered to your doorstep almost instantly, at astonishingly low prices, are among the country’s top draws. Yet these conveniences come with hidden costs, and China’s path to becoming an e-commerce giant has been far from certain.

How do digital platforms from Alibaba to Didi navigate shifting politics, and how have central and local governments responded to these ventures? Can internet-based businesses liberate and empower, or is the freedom to buy and sell ultimately a form of entrapment? For this episode, Yangyang spoke with business school professor Lizhi Liu and science and technology studies scholar Jack Linzhou Xing about the distinctive trajectories of China’s e-commerce giants, and the promises and pitfalls of the digital economy.

Guest Bios:

Lizhi Liu is an Assistant Professor at Georgetown University’s McDonough School of Business and a faculty affiliate of the Department of Government. Her research examines the intersection of politics, trade, and technology. She has received multiple research awards and grants across the fields of political science, economics, and strategic management. She is the author of the award-winning book From Click to Boom: The Political Economy of E-Commerce in China. She has also been named one of Poets&Quants’ Top 50 Undergraduate Business Professors and serves on the World Economic Forum’s Global Future Council on Trade and Investment.

Jack Linzhou Xing is an Assistant Professor in the Division of Public Policy at the Hong Kong University of Science and Technology. His research spans the sociology of technology and work, science and technology studies (STS), and technology and innovation policy in China, focusing on the social implications and governance of the platform economy, digital infrastructure, and digital labor. Drawing on longitudinal ethnographic research, his dissertation traces and analyzes how digital platforms, gig workers, and tech workers interact with one another and navigate a contingent regulatory environment amid China’s compressed socioeconomic transition, geopolitical tensions, and pervasive techno-nationalism and techno-developmentalism.

Related Materials:

Liu, Lizhi. 2024. From Click to Boom: The Political Economy of E-Commerce in China. Princeton, NJ: Princeton University Press.

Xing, Jack Linzhou, Jun Zhang, and Gonçalo D. Santos. 2025. ‘Mythologies of Wealth in Platform Economies: The Case of the Ride-Hailing Platform Didi in China.’ Science, Technology, & Human Values, online first. doi.org/10.1177/01622439251398910.

Xing, Jack Linzhou, and Naubahar Sharif. 2025. ‘A Processual Approach to Skill Changes in Digital Automation: The Case of the Platform Economy in the Service Sector.’ Research Policy, 54(4).

Full Episode Transcript:

Yangyang Cheng (00:00)

In the first 18 months of the COVID-19 pandemic, I was living in Chicago and left my apartment building a total of six times. I felt guilty about my privilege, but that did not stop me from exercising it. My existence was quite literally sustained by e-commerce platforms and delivery workers who brought food, necessities, as well as a few items of indulgence to my door.

Life before e-commerce can feel like a foreign country. I remember my first online order 20 years ago, when I was a college student in China. My birth country did not invent e-commerce, but it has certainly changed the game, and the digital economy has transformed Chinese society, reconfiguring relationships between individuals and businesses, between the state and the market. Despite our frequent interactions with it, an e-commerce or service platform can appear opaque and abstract.

I am so excited to have two leading scholars with us today to shed light on a vital part of our lives and of society. First up, we have Dr. Lizhi Liu, an assistant professor in the School of Business at Georgetown University and author of the fascinating new book, From Click to Boom: The Political Economy of E-commerce in China. Lizhi, thank you so much for joining us.

Lizhi Liu (01:21)

Thank you so much for having me.

Yangyang Cheng (01:23)

China is an indisputable giant in e-commerce today. In your book, you actually discuss how this was not presumed or inevitable. In fact, China’s rise in e-commerce has defied conventional wisdom. So what was the conventional wisdom? What were the assumed premises of what a country needs to have in order to build a successful e-commerce or digital economy? And how does China’s case differ?

Lizhi Liu (01:49)

First of all, I would like to put China’s e-commerce boom into some perspective, because I think for a general Western audience, it is easy to underestimate how important it has become in China. So China now accounts for about 50% of the global online retail sales, given that China only accounts for about 13% of global consumption. So that is an absolutely amazing number.

And also, I want to remind the audience that in early 2000, China only accounted for about 1% of the global online retail sales. So this is absolutely a market boom. And also e-commerce is not just a sales channel in China. It is a lifestyle, and not just for young generations, but across generations. I even heard people joking that the largest party in China is not the Communist Party, but the “hands-chopping party,” or duoshoudang.

This is a darkly humorous term for people in China calling online shopping addicts who keep buying things, even while joking that they should chop off their own hands to stop. But this success was not really expected 20 years ago. In the early 2000s, as Yangyang mentioned, actually many observers were deeply pessimistic about China’s online market, because conventional wisdom suggested that e-commerce development should be much more prominent in the Western markets, where consumers generally have higher income, and also the infrastructure is better because of higher internet penetration.

But most importantly, there is better rule of law in the Western market and higher credit card usage. There are better market institutions to enforce contracts and prevent online fraud. Why is this important? The core is a trust issue, because it’s much harder to establish trust in online transactions than brick-and-mortar transactions, because online transactions are not face to face.

Consumers have to pay and wait for a long time before inspecting or receiving the product. And also a lot of online sellers are mostly small and anonymous sellers selling little-known brands. So this is the exchange between strangers, that rarely happened in traditional societies, because people tend to trade with their friends. So in the West, because there’s a better rule of law and fewer counterfeits, and also higher credit card usage, if you get cheated online, you can always get your money back.

But in China, a lot of these favorable conditions were not there. In early 2000, there was a weaker rule of law in the markets provided by the government, lots of counterfeits, and credit card usage was lower than 10%. And so I think at the time, people wouldn’t expect that China would have a market boom.

But in my book, I found that it’s quite interesting, because Chinese e-commerce didn’t happen because China had strong market institutions in advance provided by the government. It is precisely because of the weaker rule of law that they propelled private companies, especially platform companies, to help supply contract enforcement institutions that are privately supplied, companies like Alibaba.

They create private institutional solutions, which I probably will touch upon later. For example, Alipay provided an escrow system in the payment system, and also a bunch of very strong institutions to enforce contracts in a weaker-rule-of-law environment. And this mechanism makes it possible for strangers to transact online, even when public institutions were incomplete. But what is also very counterintuitive is that the Chinese government, many people know that China had very stringent regulation in the online world, especially on social media.

So a lot of people assume that China had a lot of control over its internet space. That is only half true, because with regard to online speech and social media platforms, that’s absolutely right. But in the aspects of digital economy, and especially the economic aspect of the internet, the Chinese government was actually quite lenient and had very lax regulation in that regard, partly because China wanted to promote online commerce, to foster grassroots entrepreneurship and others.

So for about 10, 20 years, from early 2000 to 2020, the Chinese government actually practiced what I call strategic non-regulation. So this is the government trying to delay regulation, avoiding very stringent regulation in this space, even though they knew all the problems and they had the capacity to regulate the internet space. They gave platform companies a lot of room for experiments, for new institutions, and also gave them room to exercise regulations, private governance in the market, even though the whole online commerce transaction was pretty disruptive to the Chinese economy.

But they tolerated this great experiment, great market, and allowed them to grow. So, in effect, the government actually outsourced a lot of the market-building functions to the platforms. So the Chinese e-commerce boom is counterintuitive in the sense that it challenged the idea that strong government-provided rule of law must always come before market development.

In fact, there is a co-evolution between market development and institutional building. And this institution did not come from the government-provided one, but was first driven by the private companies, which a lot of people think that in our region you wouldn’t expect them to play such an important role. But this is what’s going on with China’s e-commerce development.

Yangyang Cheng (07:54)

This is so fascinating. The e-commerce development in China did not just defy conventional wisdom about how a digital economy grows, but also it defies the caricatures or the stereotypes of how an authoritarian government governs. It’s not just a monolith that tries to oppress everything. And since you mentioned Alibaba, sometimes with this US or Western-centric narrative, companies in China or outside of the US are defined by US companies as the benchmark, like Baidu becomes China’s Google, or Weibo becomes China’s Twitter, even though they are quite different.

And so one question I could phrase is, is Alibaba China’s Amazon? How are these two companies similar, and how do they differ in importance? What are the key differences?

Lizhi Liu (08:38)

Yeah, this is a great question. In fact, Alibaba’s platforms are actually more of an eBay model. So it’s not a retailer itself, but instead it provides a marketplace for third-party sellers and buyers to trade with each other. They also do not have a very large logistics or delivery service, as Amazon had. In fact, not just Alibaba, but also, for example, Pinduoduo, which is the parent company of Temu, they also follow a marketplace model. In other words, they’re not the retailer themselves.

But on the other hand, I feel that even though Alibaba’s model is closer to the eBay model, it has grown to be much more than eBay in China. There are several differences that I want to emphasize. One is following up on my discussion about online institutions. First is there’s simply much more trust-building effort, or trust-building institutions, that you can see on Chinese platforms compared to the Western ones.

For example, eBay actually had a big business in China in early 2000 and had a lot of hope to build an e-commerce empire in China. But it did not succeed, partly because it introduced the Western payment system into China, which is PayPal. But what PayPal did is they would send the payment directly from the seller to the buyer once you place an order. But the problem is that this kind of setting, or payment system, was sufficient in the Western setting, because a lot of people would pay with credit card. And if there’s a problem, you can always get your money back.

But in China, most people do not have credit card usage. What they do is that they basically do not trust the system, and they worry that they wouldn’t get their money back. So what really took off for Alibaba is because they invented Alipay, which is PayPal plus a compulsory escrow system, in the sense that if you order something online, the money will not directly go to the seller, but instead go to a third-party escrow system, which is called Alipay.

And the money will be kept in escrow until you receive the product from the seller and you’re happy with the product. So this additional layer of trust building was very, very important for the Chinese platforms. And there are other sorts of systems, such as they also allow instant messaging systems between sellers and buyers, because they believe that this will help build trust. Even though in eBay, at the beginning, they avoided this kind of direct communication between sellers and buyers, because they worried that the two parties would strike a deal without the platform.

So I think the reason why trust building is so important in the Chinese setting is because China, as I said, has a weaker rule of law and lots of counterfeits. You really need this additional layer to build trust. Otherwise, people wouldn’t trust online transactions. And this is how the Chinese platforms, in order to sustain their business, had to do additional trust-building work, not because they have better tech, but because of the underlying legal system in China and the institutional environment in China.

The second thing is that what I also found fascinating is that Chinese platforms actually had more semi-democratizing reforms than Western platforms. Because on Amazon or eBay, you as a user wouldn’t be able to really vote for rule changes, you wouldn’t be able to adjudicate user disputes. But these are very common institutional settings on Chinese platforms.

For example, Alibaba has this online jury setting, which is widely used in China’s different kinds of online platforms. So, for example, if a seller and buyer had a dispute, they can submit the case not just to a platform employee to adjudicate, but the system can randomly draw 13 jurors to adjudicate disputes between sellers and buyers. And ordinary users can actually adjudicate these cases, which is sort of a source of justice on Chinese platforms.

They also allow users to sometimes vote for rule changes, not for all rules, but they actually have a “House of Representatives” for rules to accommodate user requests. That’s quite interesting, because in the US or Western setting, you know that you can vote in the political system; in China, you cannot. However, in the West, you cannot really vote on Amazon or online platforms. But China actually had user governance, more of a user participation in the platform environment. That’s quite interesting.

And the third part is just like Chinese platforms have way more collaboration with the Chinese government at different levels than Western platforms, which we can cover later.

Yangyang Cheng (13:52)

This is so interesting. Thank you for giving us some of the earlier history about eBay’s frustrations in China. In some ways, like its Chinese counterpart, it was not a copycat: Alibaba was able to adapt to the particular conditions, socioeconomic as well as cultural, in China, and to develop its business better than the US model was.

And in that context, I would like to come to another aspect of the digital economy, which are ride-hailing and ride-sharing businesses. And on that, I would like to bring in our second guest, Dr. Linzhou Xing, who goes by Jack, and is a new assistant professor in public policy at the Hong Kong University of Science and Technology. Well, Hong Kong is so lucky to have you, Jack.

So, Jack, much of your research has been focused on one Chinese company, which is Didi, and in a poor analogy, some might call it China’s Uber. But one thing that I’ve learned from your papers is that Didi, which was founded a couple of years after Uber, in 2012, actually had quite a different strategy when it was starting out. It did not really start with private car owners as its drivers, but by co-opting the existing taxi networks in Chinese cities.

So tell us a little bit about that. On one hand, why would a Chinese taxi driver decide to work for this tech startup rather than an established taxi company? And on the other hand, how did Didi eventually take the route of Uber and shift its focus to private car owners as its primary drivers?

Jack Xing (15:22)

Thank you so much for having me, and this is really a great opportunity, and also to be on this podcast with Lizhi, whose work I have been reading a lot, and especially recently, because I’m also writing a paper with a very similar kind of, not argument, but similar framework. So thank you very much for pointing out there’s one major big difference in the history of the development of Didi compared to that of Uber.

And as you mentioned, the most important difference lies in that Didi first co-opts, they establish taxi driver networks by attracting them with cold cash, with all kinds of gifts, and then this switch to the private-car-based ride hailing. So this is what I call so-called creative appropriation. I wanted to craft a term and try to draw on the famous Joseph Schumpeter term, creative destruction.

So basically it means that there’s a new company appropriating existing networks and resources of the incumbents and then establishing their operation routines and their algorithm, or establishing their networks through all these networks and resources. And then they leave the incumbent, or they even betray the incumbent, if you want to put it that way. So I would like to argue that this is actually not a plan-ahead strategy. As you have mentioned, lots of things are not really planned ahead. It’s not about you having a certain strategy or a structure and then you have this outcome.

It’s more like what we would call in Chinese ‘crossing the river by feeling the stones’, a more contingent thing, there are lots of contingencies there. So at the starting point in 2012, for Didi’s starters, I think the initial idea is simply to run a very innovative and digitalized kind of mobility service without a clear idea of how this can be operational and viable and profitable.

But later they found out that it went into the easiest option, which is to co-opt the existing taxi driver networks. But later they found out its taxi-based business model is not profitable, for obvious reasons. The first thing is that, as you mentioned about the payment infrastructure at that time, lots of taxi transactions are still done by cash, and it’s very difficult for you to charge commission fees.

So it means that this kind of typical Uber kind of business model is not profitable as the starting point. And also, if you’re collaborating with taxi drivers, and the taxi industry is basically controlled by the government, you cannot charge commission fees for someone who is controlled by the government. So at that time, they found that somehow the taxi-based business model is not viable.

And Uber’s entry also partly inspired them to make a switch. So they decided, around 2015 and 2016, they wanted to switch the business model. It does not mean to drop the taxi drivers, but basically they changed their strategic focus to ride hailing. And of course, later lots of taxi drivers left their conventional taxi industry and joined Didi. But that’s another story.

So you also ask about the reasons why. I think one of the reasons is the lack of private cars, assets and resources in China. I kind of forgot the exact number back in 2012 or 2013, but at that time, definitely it’s very, very incomparable to the United States, in which, I think, there would be one or two or even three cars per household.

So this is one major problem, or a bottleneck, for some companies like Didi or Uber to run a renting business even in the major cities like Beijing and Shanghai and Hangzhou. So the corresponding problem is that you don’t have the cars, and you don’t have the corresponding workforce, because lots of people do not know how to drive, and of course they’re not familiar about how to drive, and driving as work.

So there are lots of heterogeneity within the possible workers who could join this business. So these are the supply-side reasons. The other reason is, again, to mention the government. So it’s very difficult to start with a highly controversial business. In the beginning, as I’ve mentioned, the taxi industry is controlled by the local government. So how can you directly confront a local government at the starting point?

So it’s quite different from in the United States, in which a startup tech firm seems to naturally have that kind of legitimacy, that we’re doing innovative things, we’re doing new things, and the governments are really in our way, so we are the legitimate part. So in China, it’s very difficult for you to position yourself, at least initially, like that. And so I feel that all these are the reasons why Didi chose this very different strategy from Uber, although later it converged with Uber regarding private-car-based ride hailing.

And the other thing I want to mention, which is very interesting, is that not a lot of people mention this. It seems that Didi and Uber are all high-tech firms, but when they’re doing a lot of these kinds of promotion work at the starting point, they use very, very old, traditional, foot-soldier style, labor-intensive and mass promotion. When you’re doing field work and when you hear all these old drivers talking about the situation back in, for example, 2012 and 2013, maybe even later, what you hear is all about posters at gas stations, all these promotional stuff, and gas stations and restaurants, and even there are different platforms, their different promotion staff from different platforms, maybe sometimes they fight with each other. You hear all these stories about, these are all very on-the-ground things going on, which is a very sharp contrast with the image of a high-tech firm.

And another interesting point about Uber and Didi’s promotion in China is that Didi typically employs lots of temporary workers with a hierarchical organizational structure. And Uber, because it’s a kind of foreign tech firm, so it inspires lots of Chinese students studying in foreign universities, who get back, who want to have an internship experience in a tech firm. So they join Uber, sometimes even for free. And so they organize small, more egalitarian teams to help promote ride hailing. But basically what they do is really labor-intensive mass promotion. So I feel that this kind of difference in the organizational culture and the positioning of the firms are also very interesting stories to share.

Yangyang Cheng (22:41)

So this is really fascinating. One detail I really enjoyed from your papers, and you just mentioned, was how when Didi was getting started, it was trying to just go to the restaurants where a lot of taxi drivers eat out and just aggressively approach them, give them cash, drive for us. And one thing you also mentioned, and following up on what Lizhi mentioned earlier, the Chinese government is not a monolith, and often its governing strategy is quite fragmented. And touching on what you mentioned earlier, when Didi was getting started in China, and from what I understand from your papers, is that the central government was quite welcoming: this is new economy, this is digital innovation. On the other hand, while Didi was aggressively trying to appropriate established local taxi companies and networks, it ran into ambivalence or even roadblocks from local governments that have entrenched interests.

So tell us a little bit about that. And how did Didi eventually make its case and build relationships with local governments?

Jack Xing (23:39)

Thank you for that follow-up question. That was a very important one. It also echoes a lot with Lizhi’s, I can say, groundbreaking contributions in this field. So I can talk more specifically about the case of Didi. I currently develop it, divided into maybe roughly three periods. The first one, of course, as we’ve mentioned, is about the taxi-based ride-hailing stage. At that time, actually both central and local governments, you can say that they embraced Didi.

They just kept their strategy of a so-called strategic non-regulation, as Lizhi crafted. So what is going on is the central government basically treated it as a technological innovation that may create an interesting positive impact on mobility service, on economy and society. And it did not take any action. And local governments were actually happy about this, because this is helping the taxi industry.

In 2012 to 2015, I think the taxi industry was in a very difficult time in different cities in China. And the taxi industry was a very important source of operational income, taxi income, of local governments, and has lots of private personal connections with local government entities. At that time, when Didi started to help taxi drivers, help the taxi industry, local governments were happy to see Didi grow.

But afterwards, when Didi switched to private-car-based ride hailing, the central government still kept embracing it, at least rhetorically. So at that time, that was 2015, 2016, the country’s going through supply-side reform, which basically means a lot of layoffs in the traditional manufacturing sectors, lots of unemployment.

And so the central government started a campaign of mass entrepreneurship and innovation. So Didi also tried to position itself as a kind of platform for micro-entrepreneurship and innovation. So at that time, central government actually embraced Didi, whereas the local government started to feel that, okay, you betrayed the taxi industry. You’re not only not helping us, but also kind of creating trouble for an industry that is very deeply connected with the local governments.

And when you adopted this private-car-based ride hailing, it means that the car supply could become a problem, and mobility within the city could become a problem. So it creates a lot of trouble for the local government. So what ended up happening was that in around 2016 and 2017, the central government basically had an official announcement, an official temporary so-called interim measure, to kind of legalize ride hailing and granted the autonomy to local governments to create their own regulations.

Whereas local government was very hostile in terms of regulations. They have very stringent regulations on the technical details of the cars. You cannot even imagine that there are regulations on cars like length, height and width, which is very ridiculous considering what kinds of business they’re running, and their eligibility of the drivers. As everybody knows, in China, like hukou and residential permit can be a very useful tool to exclude people from doing certain businesses, and also their unfair car depreciation terms.

And the ride-hailing private cars were treated the same as taxi cabs. They know that taxi cabs usually run like 20 hours per day, whereas private ride-hailing cars at that time typically run less than eight hours per day. So that means that if you treat depreciation terms as the same, it means it’s very unfair to the private cars.

And also there are lots of everyday harassment and sting operations from the street-level bureaucrats, so to speak, with the ordinary drivers. And Didi even started compensating drivers for fines committed by them. And that was a very, very messy situation, basically around the time of 2017, 2018, and even it’s extended to 2019 and even later.

So at that time, Didi started to think about, this cannot be the case, this cannot be continued. So they wanted to do several strategies. So the first thing is that it did an organizational reform. That was the time when they really established a very systematic, specialized government affairs team. Before, it was like three or five people in the Beijing headquarters dealing with something related to central government.

Later they found out, we could recruit all these former civil servants working in local governments, especially those related to mobility and transportation. They recruited a lot of people locally, and all these people now have the experience of working within a government system and started to really make government affairs strategies more sophisticated. One important thing to do is that they try to sell the smart city projects to different local governments, because they have the technical capacity.

And at that time, lots of governments started to respond to the call from the central government about, we want to do digitalization in different aspects of city administration. So smart city becomes a very important leverage for Didi to kind of seek collaboration with the local government. And of course, the major issue lies in the eligibility of drivers and cars and security measures.

So lots of compliance-related indicators are incorporated into the order-dispatching algorithms of Didi in different cities. But of course, here what I would like to make sure is that this does not mean that non-compliant drivers are driven out of the business. So it is more about priority: if you’re eligible, your car is eligible, you can enjoy a better order, you can receive orders quickly. But it does not mean that if you’re non-compliant, you cannot do the business.

So Didi people told me that no, this so-called non-compliant transport capacity is very important, especially during the peak hours. So Didi really has the incentive to, on the one hand, follow the local government, but on the other hand circumvent all kinds of regulations. So Didi kind of maintains this very precarious balance and precarious kind of coexistence relationship, a collaboration relationship, with local government during that time, and kind of embedded deeper and deeper and finally became a local infrastructure in mobility and employment in different cities.

And so it is not to say the local government finally embraced ride hailing or not. It’s more about they kind of accepted the fact. It’s already a fact there, and they have nothing to do with it. They cannot get rid of the platform, and they do not have the incentive to do that, because Didi’s business model maturing and security measures maturing, the disruptions are now becoming the so-called new normal. So finally local government gradually accepted the situation.

Yangyang Cheng (32:08)

This is so fascinating. It is contrary to a lot of the stereotypes of an authoritarian government suffocating innovation. Actually, sometimes political oppression and regulatory crackdowns can create the incentives for clever maneuvering, and can be a ground for innovation itself. And so Lizhi coming back to you and your groundbreaking scholarship, and as Jack also mentioned, some of that is in line, as you also mentioned earlier, with how local and central governments have outsourced some of their governing functions to these tech platforms.

And for e-commerce, everything from payment systems to consumer disputes. And for a while there was a strategy of strategic non-regulation, or a relatively cozy relationship between Beijing and these new tech firms. But that relationship did not last forever. And starting in, basically, late 2020, the Chinese government has initiated a series of crackdowns on these high-profile new digital economy companies, including Alibaba and including Didi, especially with their IPO attempts.

And so can you tell us a little bit about that? Why did the central government suddenly turn on these tech giants? One narrative we hear in the US is it’s ideological. Xi Jinping is a dictator, it’s the Communist Party, of course it’s going to crack down on private enterprise. But is it that simple, or are there specific practical concerns as well? And were there specific triggers that led to this wave of crackdowns in the early 2020s?

Lizhi Liu (33:48)

Yeah. I think the regulatory crackdown was driven surely by both political concerns and more concrete economic rational concerns, even though I think Western media tended to cover more the political grounds. But I would argue that this kind of ideological concern with private capital has always existed in China. And people have written about, for example, starting from mid-2010 or even earlier, there’s already an ideological turn and more tendency to regulate the private sector.

But as we know, the crackdown on tech companies actually kicked in in late 2020. So I would say there’s a time inconsistency there, showing that even though there’s a more general ideological turn, still, there are some other reasons behind this regulatory crackdown. So let me just talk about several things that the government really wanted to fix with more stringent regulation.

One thing is that this is actually a global trend of regulating tech companies. Let’s say Europe, or even the US, in fact, started to regulate private tech even before China. And actually in China, people would say, especially in the policy circle, they would say they’re just simply following what the US has been doing. The only difference is that the Chinese government was way more powerful than Western governments.

And in the West, if you want to do antitrust regulation, for example, the case between the US government versus Microsoft, the antitrust case actually lasted for over a decade, and in the end the case was resolved. But in China, when the Chinese government really wanted to, for example, regulate Alibaba and Meituan on similar grounds, anti-competitive cases, they did it within half a year. So that’s the difference between political systems and how far you can go with the regulation.

But I think a lot of these considerations of why the government should regulate private tech are pretty similar. Just to give you several examples. One is that surely there are some concerns about anti-competitive behavior by platforms. And so, as I mentioned, basically in the first two decades of e-commerce development, actually this is not just about e-commerce, but a lot of digital platforms, especially digital economy sectors in China in general, the Chinese government basically granted some sort of, in China people call it jianguan hongli, meaning dividends coming out of lax regulation.

In other words, because of this more or less strategic non-regulation and lax regulation, the platforms gained a lot of autonomy to grow. And we know that in the platform economy there’s a tendency of big platforms getting bigger, because of network effects and everything. Therefore big platforms started to essentially, as their power consolidated, started to manipulate their market positions and started to have certain anti-competitive behavior.

One of the common practices with big platforms is called “either-or” restriction, er xuan yi. So this is a practice that sometimes, for example, during a sales event, an online platform would have some locking deals with certain sellers, saying that you can only run the biggest discount on my platform, not on my competing platform. Otherwise you will lose this ranking, you will have fewer consumers to visit your site. So this is surely anti-competitive behavior. And that then becomes a very important aspect of regulation that the government actually did.

And the second one is over-collection and selling of personal data. In fact, in China there’s a big black market for personal data. It used to be so cheap to get personal information. There is some news about how information can be sold for as little as 2 yuan, which is $0.30 in US dollars, per individual, in 2018. So in other words, there’s a common phrase on the internet about the internet economy, saying that if you are not paying for the product, you’re the product. Because even though the internet services seem to be free, they’re not actually free, because you’re basically giving tech companies extensive access to your personal data.

And the third very interesting issue that a lot of consumers were not happy about is called dashuju shashu, big data enabled price discrimination against loyal consumers. So there is a professor who did some research using basically different types of phones, Android phone versus Apple phone, to call Didi, right, ride-hailing apps in China, and found that he’s paying an Apple tax. So basically when he’s calling Didi using his Apple phone, he got higher fares and managed fewer opportunities for discounts. So it’s very interesting how the big tech companies can actually use this to enable price discrimination and get more profits. So these are the issues that the consumers were long complaining about, even before the tech crackdown kicked in.

So I do think that even though there are definitely some sort of ideological concerns and considerations that triggered the tech crackdown, including the slogan that one of the purposes for the crackdown is to prevent the disorderly expansion of private capital, put that aside, there’s still some economic rationale to why the Chinese government wanted to regulate private tech, and we should pay attention to the latter factors as well.

Yangyang Cheng (40:32)

Oh, this is so fascinating. Thank you for bringing in, this is not just a story about companies and the state. There are also people, including the workers and the consumers, whose rights and whose lives and livelihoods are being impacted by the platform economies. And so on that note, I would like to stay with you, Lizhi. You studied 100 Chinese villages across three provinces for your book, in terms of whether or not e-commerce can deliver the promise of prosperity. So what did you find? Did participating in e-commerce make the villagers’ lives better, and in what ways?

Lizhi Liu (41:13)

Yeah. So my coauthors and I did a field experiment in 100 villages across eight counties in three provinces of China. So what we do is we try to understand whether introducing e-commerce into rural villages would actually improve rural livelihoods, and through what channel. So we basically have these 100 villages, and some of the villages were randomly selected to be connected to e-commerce channels, whereas other villages are not.

So we’re basically comparing the villages with very similar economic preconditions and population, and comparing those with e-commerce connections and those without, to find out whether introducing e-commerce would actually change the farmers’ buying and selling behavior, and whether there’s something working because of e-commerce. So what I found is that e-commerce mostly brings in this kind of consumer-side welfare, but not so much on the production side.

I will explain one aspect of consumption first. So essentially with e-commerce, the farmers can get cheaper and greater variety of products. They also mostly benefit from buying durable products, especially household appliances, because without e-commerce, in the past, it is extremely difficult for rural villages, especially those remote villages, to get very quality products.

Sometimes in order to buy household appliances, they have to travel on average 40 minutes to the township or nearby city to actually buy those expensive items. But with e-commerce, this can be shipped to their household directly. So that’s a huge gain. So in other words, e-commerce really equalizes consumer rights across urban and rural areas.

However, what we found is that it is not easy. Even though e-commerce in theory can enable both selling and buying behavior, in other words, farmers probably can also sell through e-commerce channels, we actually found that it is extremely hard to turn rural farmers into online sellers overnight. There had been a lot of promising stories about how villages in China, indeed there’s thousands of them, villages in China suddenly become an e-commerce hub, a lot of farmers started to sell through e-commerce there.

Indeed, that’s the case in China, under certain special conditions, either they have a farmer who’s very entrepreneurial, who basically led the whole village to change, or they have some suitable products offline so they can really sell online. However, what our experiment shows is that it is extremely hard for an average or normal rural area to suddenly use e-commerce to sell these products, because a lot of these agricultural products are very perishable and difficult to ship.

They’re not like clothing or cosmetics. So it’s not very easy to gain new income for the farmers. So I think that also has some implications for how we think about whether e-commerce can bring in sort of equality in income. Because at the beginning, I think a lot of people hoped that with e-commerce, in theory, everybody can sell to the national market. So for the marginalized groups, they probably will benefit from it.

I think on the one hand, it’s true, because consumption welfare is also very important, meaning that you can get the products at a cheaper price. But on the other hand, it’s not so easy for people to just take up this digital channel to suddenly become an entrepreneur and get additional income. So I think that is a very different channel, and it’s quite interesting to help us think about how e-commerce can really generate a broader socioeconomic impact.

Yangyang Cheng (45:36)

This is really interesting because, as you said, the promise of prosperity and equality may be true in some ways, but it’s actually a lot more complicated. Freedom is not just the slogan. And one thing that I also learned from your book is that, for example, the freedom to sell is a complicated matter, in terms of whether you’re free from some of the local regulations by relying on the e-commerce platform, but then you are also becoming dependent on the platform itself and some higher-level regulations.

And in some ways that also echoes driving for a local taxi company versus driving for a platform like Didi. And so Jack, coming back to you. One thing that was really interesting is you wrote about this concept of a myth. Like, why would someone drive for Didi? What is the promise of prosperity and autonomy? And what is the reality that Didi also actively destroys after appropriating the taxi networks that it had utilized to get its start, and how its algorithms actively extract the tacit knowledge, the temporal, geographical knowledge of the taxi drivers, feeds that into the algorithms, and then later basically deskills the workers who drive for its platform. So tell us a little bit about that. How does a platform like Didi engage in myth-making and deskilling of workers?

Jack Xing (47:11)

Great, thank you so much for that great question. So I can just continue with the discussion about the promise of prosperity. Of course, it’s a promise of prosperity, autonomy, and I think it’s a global thing in terms of how the platform economy portrays itself and promotes itself. And for Didi, what we found is it has a kind of very typical Chinese structure.

So if you interpret it in a more anthropological kind of way, you try to analyze the structure of the story. There are several typical elements there. The first one, surprisingly, is about risk taking, because this is about using your private assets. So it’s about using your car. So this risk-taking part is that you borrow some money from your family and you get a car.

So that’s one element. The other element is that you’re very hard working, that you drive like eight hours, ten hours per day. And this is the second one. The third one is technology. So you are really going to go into an era that is very hopeful, which is technology. It brings you basically anything related to technology, it can bring you this so-called prosperity and hope and great income, great achievement, etc.

And the other one is luck. So luck is kind of the final last resort. So if all others click but you still did not succeed, it means that you didn’t have that luck. So that’s the thing. So you can see that you have these four typical elements for a story, which actually echoes a lot of previous anthropological research just about rural migrants, about the global Chinese diaspora, how they try to climb up the class ladder, try to gather personal prosperity, try to get their family better, etc.

So I think this is not entirely a new story, apart from the technology part. That part is quite new in this story. So this is a very interesting mythology. And one particularly interesting episode I want to share is that every time when I interviewed all these drivers, especially in the earlier days, I would explain to them about how this algorithm works and how these platform economies are organized through certain kinds of business organization, and why these platforms can give out all these kinds of subsidies, because this is backed by global financing, etc. And they just become very, they just don’t want to know.

And they have no interest. So they explicitly said, I don’t understand this, and I don’t think I understand it, and if I succeed, I succeed; if I fail, it’s because of luck, or it’s because of ming. What does ming mean? Ming is about fortune. So you can see that this kind of mythology plays a much more important role in that. Technology itself is not about the content of a technology or a business organization to give people a promise. It’s more about this kind of storytelling. This is kind of a deeply cultural thing.

And interestingly, I also find some symmetry inside. So we were discussing this topic with all these professionals working in all these investment entities, like investment banks, as well as venture capital. What they told us is that, no, you don’t want to lose the next Facebook, you don’t want to lose the chance for this firm to become the next Tencent. So the next Tencent is basically the Chinese version of the next Facebook. So this kind of understanding is not really based on rationality, which we typically link with technology.

It’s really about how people understand their pasts and their futures, about understanding their life. And of course, the story structure is quite symmetrical on the side of investors and on the side of drivers, but their fortunes, at the use of the word fortune again, the result was quite different. So for the drivers, of course, when they go in, they found that, okay, so all these are not necessarily completely wrong, but more of a myth. That means that it’s not as easy as you think.

And of course there are business cycle ups and downs, and typically platform economy firms, they have a period in which they want to do this negative profit, loss-leading kind of expansion, and at that time they give out lots of subsidies. But afterwards, when markets are mature, all these subsidies are gone, and the drivers start to suffer from longer work time but shrinking income. Whereas for investors, if this fails or they succeed, they can always go to the next potential Tencent or Facebook. So I think that’s a very striking contrast between how similar their stories are and how different the outcomes could be.

And your second question is about deskilling, or what I prefer to call skill change. Because sometimes I feel that deskilling is not a very accurate term, because after I do my field work with all these drivers, I feel that actually their skills are not gone, it’s just like they’re forced to use their skills in a different way, which I call repositioning and refocusing.

So I think there are several important points about it. The first starting point is how we treat skills. So typically when you see all kinds of studies in economics and even in sociology, we kind of have a very categorical and hierarchical understanding of skills. Which skill is of a high value, which skill is of a low value.

So I think this is quite, we have already seen it in this current era of digital automation, this kind of categorical or hierarchical understanding can be very difficult to sustain. For example, if you are using generative AI, what kinds of skills are you using? I mean, is that intellectual? Is that intuition? Is that based on training, or is it really about tacit, is it really about experience? You cannot really figure it out.

So the starting point is that do not use this categorical and hierarchical understanding to look at skills in the current wave of digital automation. And if we take that as our starting point, we start to think about how exactly a skill is. It is really something about judgment and action. On one hand, you kind of make a decision on what you want to do, and then you execute it. It’s so simple. And then your work is like a process. It’s like a series of events. It’s the event after event. You do the judgment, you conduct your action, and then this work process moves on. And now you have the technology. So the skill change is basically about how the judgment and actions are changed or interrupted by this automation technology.

And so this is the second point. So I can raise one very interesting example about drunken passengers. We can compare the conventional taxi and ride hailing. So if you were a taxi driver, it basically means that you can plan, you can do a spatial-temporal plan, you can basically decide what kinds of passengers you want to take and what kinds of orders you want to take.

So if you encounter a drunken passenger, a typical taxi driver just will drive away. And of course, this is illicit, this is illegal, strictly speaking. But nobody can really enforce that kind of regulation. So you kind of avoid a situation, or avoid future events, that may cause you trouble. So this first step is about picking up passengers.

Whereas compared to a ride-hailing driver, you don’t know this is a drunken passenger. You only know at the point when you’re picking him or her up. So it happens that you have no chance to reject that, and the order is kind of matched by the algorithm. You cannot reject it in the face of that passenger, because there will be a penalty on your service star rating, and then you will be penalized for future order opportunities.

So you’re forced to take that passenger. So this is the first step. You can see that the algorithm, or the whole platform system, basically rips off your opportunity to make that decision. And if that stops here, then it means this. But actually it is not. So think about what would happen later.

So for taxi drivers, the story ends, you avoid this kind of chance to get into trouble. But for a ride-hailing driver, things need to go on. And maybe during the trip, the passenger would just vomit in your car. And then now it’s the time for you to make the judgment into action. And this event is actually created by the algorithm. It’s basically assigned by the platform algorithm to you. And the algorithm now does not help you make the judgment or action. Now it’s only up to you. For this event, you need to deal with it all by yourself. So you need to think about several things. The first thing is whether I want to send this person first to his destination or to a nearby…

Firstly, I go to a nearby car wash to clean up this mess. So that involves lots of judgment: whether you want to serve this passenger, lots of emotional labor, you want to communicate with that passenger. And also it’s about spatial-temporal knowledge you have. Do you know whether there is a car wash nearby, whether it is convenient for you to send him to his destination first, or you go to the car wash first?

And so now it is the situation in which, if you have the experience of being a taxi driver, your skill can still work. Because typically if you have experience with driving taxi, you’re very familiar with the situation nearby. Whereas if you are a fresh ride-hailing driver, you might not necessarily know the situation nearby, because you lack the kind of spatial-temporal knowledge about the city.

So you see that your emotional skills are the focus now. On one hand, the other thing is that spatial-temporal skills, which are central to the conventional taxi, are now marginalized, but it is still relevant. So this is what I call repositioning. Looking at the whole work process, previously it is about you using your skill to control and manage the spatial-temporal work and plan the whole working process. But now it becomes, you are actually addressing the extra and sometimes unpredictable events created by the algorithmic judgment. So it means that you become someone who maintains and fixes the algorithmic decision so that you can make the algorithmic system work. So previously you’re at the center, now you’re at the periphery. You are the kind of secondary role to help this whole system of work.

But this secondary role does not mean that you are not important. You’re still important, you’re still essential to the process. But people, or the tech firms, or the policy makers, or the passengers will think that, okay, you’re at this deskilled, you’re just doing a very low-skill, low-value job. But actually, in terms of importance regarding finishing the job, I think it’s equivalent to taxi drivers. So this is why I prefer to call it skill change. And I think the concepts of reskilling and deskilling are kind of problematic. So just to share this interesting typical example of drunken passengers.

Yangyang Cheng (1:00:14)

Oh, thank you so much, Jack. And listeners, I actually do not know how to drive. So a company like Uber or Didi will not be able to extract or appropriate a skill I do not have, and I will refrain from commenting on whether or not I have been a drunk passenger. But jokes aside, I really appreciate all this context that you just gave, because the worker whose labor is exploited and whose skills may be appropriated and extracted by algorithms are often rendered faceless and voiceless in this platform economy.

However, there is one Chinese writer who is a delivery worker himself, who has written this rare first-person account of surviving in China’s platform economy. I Deliver Parcels in Beijing has been translated into English by Jack Hargreaves, and I do highly recommend it. And so Jack, not the Jack who translated I Deliver Parcels in Beijing, I know that you’re also familiar with the book. And there was one instance in Hu Anyan’s account of being a delivery driver that really struck me, was when the delivery company he was working for was going out of business. And so in the final few months, there were no quotas to fail, there was much less pressure, and he actually enjoyed the work, because he could take his time, he could familiarize himself with the neighborhoods.

He could even interact with his customers and clients. And of course, that was under a very specific scenario, and that was, in effect, a dead-end scenario. However, I was curious whether or not that actually points to openings to alternative futures, where delivery workers, whether it’s drivers or delivery workers or people who work in e-commerce, they perform an essential service and their work is essential. And so are there ways for that work to be fulfilling and dignified?

Jack Xing (1:02:13)

So firstly, the book is a great one, and we have already seen a lot of other similar books with similar contents, and we’re always calling for more and more, because I personally think that it is the first step for all the public, for the readers, even if they’re casual readers, to understand what is going on with the platform economy, what is really going on on the ground for digital automation.

So without that kind of knowledge, our discussion will be only limited to a certain level, in which we talk about, okay, so AI is displacing jobs, whether this job could be displaced by this kind of machine or not. So these are very dichotomous and superficial discussions, which really do not catch the core of the issue.

So, and thank you very much also for this good question. And a very, very difficult question. So I of course do not have an answer to what are the better alternatives. But I do want to point out several important points for us to make it possible to think about maybe better futures, in which all these workers who are in a position of helping an autonomous system or autonomous machine can gain more welfare, or at least gain more dignity and more understanding from the society.

I think the first starting point is the issue of design ethics or moral imagination. What I hear a lot when people are discussing whether there are better alternatives, what I usually hear about is that, okay, all these platforms are already doing their best. There is an equilibrium between the platforms, the passengers or customers, and the workers. And if the system changes, one party will lose out and the equilibrium will be suboptimal. So now it is the best situation. And this is a very tricky rhetorical strategy or discourse that we are hearing a lot.

So they firstly acknowledge the limitation and perform a kind of goodwill. So by acknowledging the problem, that spreads sympathy, the speaker positions themselves on the same side as the critics. And the second step is that it performs a kind of helplessness, by claiming that every alternative has been considered and none works better, the speaker makes the status quo appear to be the only possible outcome. And the third one is a kind of shutdown of imagination. This is very straightforward: if you claim that there is no better alternative, of course you will not think about alternatives.

So I think this ethical position or moral position is very important. Everybody should think about, never say that a certain party has already done the best and there’s no better alternative. This is kind of a way to escape the moral responsibility for the harms that systems produce. Whoever says that is actually benefiting from the system and he or she does not want to change.

And there’s a power-related issue. If you can claim that, it means that you have some superiority within that power relationship. So I think that is very abstract, but I think it’s a very relevant moral position to take. Like when we’re talking about whether these ride-hailing drivers or food delivery workers, they deserve better social welfare, they deserve social security, they deserve a platform to invest in their safety, etc. Just never say that, okay, the platform designers have done our best. So this is the wrong moral position to take. So this is the first point I want to make.

The second point I want to make is that I personally think that the current discussion about how to protect workers from the impact of automation and platform economy has largely missed a point. So around the globe, including China, there are basically two paradigms of thought about how policies and how society need to respond to that. The first one is basically called training or upskilling. It is about saying that, okay, all these people are low-skilled workers, so we could train them, we can make them more knowledgeable, we should upskill them so that they can possibly do a better, more valued job.

And my question to all these people is, what kind of better job are you thinking about? Is that staying in a tech firm? Is that being employed by OpenAI and creating AI systems? Is that good enough? And all these people are also being laid off. So what is the end of upskilling? The other one is about simply compensation. And the extreme version is, of course, universal basic income. But this understanding is very popular in Europe, and lots of people are calling for it even in China. My question to them is, again, what is the end? So could that be happening, and is that compatible with our current political economic situation in the globe?

So what I think both paradigms miss is that it kind of downplays the role played by all these workers who are in between. So they are really maintaining the algorithmic system, the automation system, and they are making the system work. For example, we’re using AI to, for example, write media articles, write things. We cannot stop by letting the generative AI generate something, and that’s the final product. We still need to check, we still need to correct, we still need to reiterate, to use the term that tech firm people want to use.

So in that sense, our role is very secondary, but they are very essential to make the system work. So all these people, they cannot be completely unemployed. They also cannot be the master of the technological system. They’re in between, but they’re very important. So for the society, for the policy makers, I think it is very important to focus on this kind of people, food delivery workers, ride-hailing drivers, of course these are the older version. Now we also have data annotators who annotate data to feed into the AI system. We also have all these jobs typically involved in using and interacting with generative AI. So how do we treat all these people? How do we understand their value in the whole work process?

I think that is a starting point to rethink skill, rethink the value of work, not by hierarchical terms and not by whose educational level is best, but by their essentialness. So I think that is a very important starting point to really imagine and think about social welfare policy tools, public attitudes towards the relationship between labor and technology, knowledge and labor and automation.

This is the second point. And a third point I want to make very briefly is that when we criticize platforms, we need to understand this whole problem about labor rights protection, the landscape, the overall very poor landscape, in places like China. So all these workers choose gig work because they think it is already relatively better. So perhaps besides criticizing the platform economy, we need to have a kind of overall rethinking, critical thinking about the whole labor rights regime in China. So that’s the three points I want to make.

Yangyang Cheng (1:10:37)

And this is an excellent place to tease our upcoming episode, which is about labor organizing and class solidarity in China and beyond. And I think even when our lives and our labor may seem to be captured by tech platforms, with moral and ethical imagination and with collective action, that can open up escape hatches to alternative futures. And on that, Dr. Jack Linzhou Xing, thank you so much for joining us today, and all the very best with your new position in Hong Kong.

Jack Xing (1:11:08)

Thank you so much.

Yangyang Cheng (1:11:09)

And Dr. Lizhi Liu, thank you for sharing your groundbreaking scholarship. And congratulations again on the fascinating new book.

Lizhi Liu (1:11:16)

Thank you so much, and thanks so much for the wonderful discussion and fascinating work. Jack, I definitely learned a lot and highly recommend his work to all of our listeners.

Yangyang Cheng (1:11:28)

Absolutely. And we will have links to our guests’ published works in the show notes.

The post Episode 12 | Entrapped by E-Commerce appeared first on Made in China Journal.