Skip to content

What to do with Google—nothing, break it up, nationalise it, turn it into a public utility, treat it as a public space, or something else?

Google has become a service that one cannot go without if one wants to be a well-adapted participant in society. For many, Google is the single most important source of information. Yet people do not have any understanding of the way Google individually curates contents for its users. Its algorithms are secret. For the past year, and as a result of the European Court of Justice’s ruling on the right to be forgotten, Google has been deciding which URLs to delist from its search results on the basis of personal information being “inaccurate, inadequate or no longer relevant.” The search engine has reported that it has received over 250,000 individual requests concerning 1 million URLs in the past year, and that it has delisted around 40% of the URLs that it has reviewed. As was made apparent in a recent open letter from 80 academics urging Google for more transparency, the criteria being used to make these decisions are also secret. We have no idea about what sort of information typically gets delisted, and in what countries. The academics signing the letter point out how Google has been charged with the task of balancing privacy and access to information, thereby shaping public discourse, without facing any kind of public scrutiny. Google rules over us but we have no knowledge of what the rules are.

One might think that, if the problem is that Google has become too powerful, the solution might be to break it up as a monopoly so that instead of having one dominant search engine, we end up with multiple smaller search engines. This alternative, however, is not without problems. It is easier to regulate one search engine than to regulate dozens of them. If one is concerned about privacy and the right to be forgotten, for example, having one dominant search engine (or two or three) is likely to be a more favourable option for exercising better control. Having too many search engines may also carry inconveniences for users. Part of what makes Google so useful is the large number of users it has, which improves its data and makes its tools more accurate.

If breaking up Google is not desirable, what are the options? Decades ago, one might have considered nationalisation. Today, however, that does not seem like a realistic option, among many other reasons, because Google is an international business, and it could not be nationalised by a single country or political union. Perhaps more importantly, it would probably be unfair to Google. A business should not be punished for its success.

It can be argued, however, that an institution with so much power and which profoundly shapes the lives of a good portion of the world population should not be a business that, by definition, has profit as its primary concern. In order to secure the interests of the public, companies like Google must be better regulated.

One option is to turn Google into a public utility. Adam Thierer has argued, however, that Google cannot be a public utility for two main reasons: 1. It does not possess the potential to become a natural monopoly. That is, the hottest networks constantly change, and Google could quickly become a thing of the past. Mark Jamison recounts how, when Google search results contained a noticeable error on the morning of January 31, 2009, users simply switched to Yahoo and other search engines. 2. It is not an essential facility. That is, if Google were to suddenly disappear, the economy would not shut down as it would if electricity was lost. Furthermore, one of the most important justifications to regulate public utilities is to guarantee reasonable prices, but Google is a “free” service. This last point is debatable, since Google makes users sell their data in exchange for its services, and it can be argued that the amount and type of personal data being sold is an unreasonable price to pay.

A further option is to treat Google as a privately owned public space. After all, Google does not provide a physical service like electricity or water. Rather, it has created a (virtual) space where people can search for whatever interests them. At least in the United Kingdom, however, legislation of privately owned public spaces is controversial. Private owners can restrict or refuse entry to members of the public—a policy that does not exactly protect public interest. Moreover, Google is not just any public space. Google is the contemporary version of the ancient agora. Together with Facebook, it is the central spot in the international city of the Internet. It is where people go to find each other to exchange ideas, information, and goods.

Perhaps none of the options explored here are good enough, and maybe this is a sign that we cannot fall back on old formulas for regulation. We need new international laws that adequately reflect the realities of our time. We need regulation that acknowledges giants of the Internet such as Google as the virtual agoras they have become, ensures that public interest is satisfactorily represented and protected, and at the same time does not punish Google as a business for its success. It’s a tall order, but we should settle for nothing less.


Share on

15 Comment on this post

  1. Some really quick thoughts.
    1) Electricity going down would be the same as all internet search going down, which would effectively bring us back to the internet as it were before the mid 90s, which I imagine would, in fact, have a pretty big effect in the economy (albeit still orders of magnitude less than losing electricity). In this case, Google going down would be close to a huge electricity provider going down. We can consider search engines some sort of international semi-public good. But this leads to the second point.
    2) We seem to lack an international entity capable of regulating such things, and it does not seem we are any close of having one.
    3) I worry the level of integration Google and – more acutely – Facebook have is underestimated. I believe they can be seen as forming extended minds or identities, and their algorithms, rankings, timelines, etc. are in a sense integral parts of peoples’ selves. In this view, people would be to some extent an integral part of the whole algorithm. Then, some of your worries disappear. On the other hand, they reappear as the problem of corporations having goals in conflict with the goals of the individuals whose selves are integrated with its products. Breaking the monopoly and allowing a multiplicity of search engines would help here, but would also – as you mentioned – make regulation harder. However, not only I am convinced we lack, and will continue to lack for some time, any sort of international regulatory force here, but also I am convinced that getting a good international regulatory force will take even longer. Thus, it seems we have reasons to prefer a free-market solution for now.

    1. Thanks for your comment, Joao.
      1) You are right. But in some countries there is only one electricity provider, because electricity providers are what are called “natural monopolies.” So if that one electricity provider goes down, the economy of the country shuts down in a way that does not happen with Google. But if we think about it internationally, Google does look more like a public utility.
      2) Even if you accept that G and FB are extensions of selves, my worry persists. If anything, it gets worse. FB may well be an extension of myself, but I have no idea what it’s doing to my information—or rather, what it’s doing to myself. The problem of international regulation is a huge one, of course. But I am not convinced that, therefore, the best option for now is to stay put. Then things will never change for the better. The right to be forgotten, for all its problems, is a proof that regulation, at least at the level of the European Union, is possible.

  2. Hmm, so Google builds an incredibly useful product, and supplies it for free, and your response is to try to take it away from them. How many of those 80 academics have produced anything half as useful as Google? (oh, but you don’t want to “unfairly punish them for their success”, apparently “fairly” punishing them is okay with you.).

    Have you considered the possibility that the exact reason that Google has been so useful and such a great contributor to the public good is because of a lack of regulation? It is noticeable that the most innovative area of technology in recent decades has been IT, which is the least regulated.

    1. Thank you for your comment, Tracy. You are right to call me out on claiming that I do not want Google to be “unfairly” punished. I do not think they should be “fairly” punished for their success either (it would not actually be fair); they should not be punished at all. Businesses should be punished for bad practices, but not for their success. That was a mistake on my part. And no, I do not want to “take it away from them.”
      Yes, you are right that in many cases lack of regulation is good for innovation. That is not always the case, however. Sometimes regulation of monopolies is precisely what enables smaller competitors (who innovate) to emerge.
      Also, it’s worth having in mind that regulation need not always be burdensome for businesses. I happen to know people who work for Google who think that, although the right to be forgotten is a good policy to have, Google should not be making those kinds of decisions—they should be made by someone whose priority is the public interest. A regulation that puts that burden elsewhere would likely be welcomed by Google. It would certainly save them money, effort, and problems. Having more regulation in the sense of having public actors supervise or collaborate in certain decisions with Google (e.g. designing algorithms), or requiring more transparency of Google, might likewise be good business for them. It would give them greater legitimacy. As a result, people would likely trust them more and use their services more.
      In any case, I stand by my main point: I do not want a private business tailoring what we can see on the Internet through secret algorithms that face no accountability. Too much power almost always (if not always) ends up in abuse. I do not want a private business keeping so much personal information about people without having to follow any ethical guidelines. The stakes are too high.

      1. Carissa wrote: ” Sometimes regulation of monopolies is precisely what enables smaller competitors (who innovate) to emerge.”

        It depends on the monopoly. In the case of the internet you could argue that there are technological monopolies and natural monopolies. Technological monopolies should (in theory) not persist, given the strong incentives to innovate. Natural monopolies (ie dominant firms in markets where there are gains from scale), such as utlities companies, often face price regulations (to reduce rent-seeking). But in Google’s case the price they charge users is zero.

        “Also, it’s worth having in mind that regulation need not always be burdensome for businesses.”

        Ever worked in the private sector? 🙂 Real markets are regulated. The idea of a truly free market is a bit of a myth (or perhaps a simplification, or idealisation). In the case you’re talking about, it’s disingenuous to argue that the regulation would not be burdensome for Google. The whole point of your original post is to argue that Google should be hobbled because you think it’s too big, or too all-knowing. Basically I agree with Tracy’s points – the effort to make Google bow before the European Commission looks like an exercise in the politics of resentment. Personally I don’t see the problem with Google. Alternative providers exist, and as long as Google doesn’t start eating them or killing them, then people who are uncomfortable with Google can simply use another provider. Making that switch is zero cost, so it’s hard to see how it’s a big deal. Personally I don’t understand the continental preference for killing successful companies or insisting that anything successful is public property.

        1. Thank you for your comment, Dave.

          On switching to another provider: 1. People may not be aware of some of the problems with Google, so they may not be motivated to switch. E.g. Consider the antitrust accusations being held against Google that the company distorts web search results to favor its own shopping service. If people have no way of knowing about this (assuming, for the sake of argument, that the accusations are true), they will not use another provider. 2. It may not be so easy to opt-out (e.g. Google owns YouTube, and there are no equally good alternatives to YouTube) 3. Switiching does come at a cost—it is quite likely that for many kinds of searches Google is more accurate because of its size and the amount of data it has.

          On regulation: You are right in pointing out that a truly free market does not really exist. So perhaps the more precise way to frame this issue is not whether to regulate or not, but rather, how to regulate (what is the best regulation to protect and be fair to all stake holders?). Yes, I think we need more oversight, but I’m not convinced that such oversight will necessarily be such a terrible thing for Google. (I reject the options that would be more dramatinc: breaking it down, nationalising it, or turning it into a public utility.)

          Consider the following hypothetical regulation: Google must show an independently appointed ombudsman (or a similar figure) its algorithms. Let’s suppose, for the sake of argument, that Google can be guaranteed absolute confidentiality so that nothing can get leaked to its competitors. Would it really be such a terrible thing? Would it “hobble” Google, “take away” what’s theirs, impede innovation, “kill” the company? I don’t think so.

          Consider another hypothetical regulation: Google should not be in charge of requests concerning the right to be forgotten; public officials should deal with that. Would it be so terrible? Google would be happy with this proposal, no doubt about it. They certainly did not want the burden on their shoulders in the first place, and they have had to hire a whole team of people to deal with this huge challenge (which I’m sure costs them a lot of money). There are problems with this proposal, of course. It is not clear who in the EU could have the necessary team to process the thousands of requests. But that is not the point. The point is that regulation to oversee some aspects of Google that can endanger public interest does not have to (and should not) amount to destroying Google.

          1. Hi Carissa, thanks for replying.

            Your first point (1) is one about consumer knowledge (or lack of knowledge) rather than company practice per se. I like voluntary initiatives, labelling requirements, etc in which companies are clear and transparent about what they are doing. (Voluntary initiatives are sometimes less distortionary than regulatory imposition.) But I don’t see why Google should be required to do more than make the generic details of the information it stores on individuals available, so that they can see whether they’re comfortable with that information-gathering. Point (2) is pretty minor. I struggle to see why legislators should get out of bed so that someone can see a dancing cat on a non-Google provider. And point (3) strikes me as just as trivial – so it’s a little inefficient not to use Google. It’s often inefficient to avoid companies when one is in a huff with them. There are airlines I won’t use, and clothing companies and pizza firms, and both these individual choices occasionally have very minor costs associated with them. But that’s not a reason for the state to endorse my preferences.

            I don’t see the case for the Ombudsman, either. Unless human health is at risk we don’t usually let the state simply prize open valuable IP and poke around inside it. And I don’t see how knowledge of that IP would address the concerns you have, which are about privacy. Knowing what the algorithm is wouldn’t address that (unless you made it clear to the public how to avoid detection by that algorithm, which would seem to violate the point you made about the Ombudsman working in confidence).

            In terms of Google being in charge or not of requests to be forgotten – I’m not sure such a right exists. And I think it’s naive to think if such a right does exist that a single company should be expected to enforce it. Making Google forget people without similar requirements on bing or altavista or yahoo (etc) strikes me as giving those firms excellent incentives to specialise in remembering people who want to be forgotten.

            I think what you really want is summed up here: ” I do not want a private business tailoring what we can see on the Internet through secret algorithms that face no accountability. Too much power almost always (if not always) ends up in abuse. I do not want a private business keeping so much personal information about people without having to follow any ethical guidelines.”

            The answer is simple. Don’t use Google. Tell your friends. But nothing you’ve yet written makes a compelling case that the state should take your side.

            1. Thanks for your reply, Dave.

              If I understand your view, you think that regulation:

              1. Would be disastrous to Google
              2. Is unnecessary (and in so far as it is unnecessary, it is unjustified)

              Regarding 1: I tried to give some examples of kinds of regulation that would make things better and not be disastrous to the company. The examples were not meant to be exhaustive. Further regulation would be needed to protect privacy, but we can imagine similar non-dramatic laws to take care of that (e.g. requiring safety procedures when it comes to personal data, etc.)

              Regarding 2: We probably have different views about what is the role of governments and are unlikely to come to an agreement here. Just briefly, however, I don’t think we can compare Google to a pizza company or an airline. On the possibility of an Ombudsman: Newspapers have ombudsman who represent the interests of the people, whose concern is not making the company richer, but democracy and fairness. This role is important and works well. People’s health need not be involved, as in the case of newspapers. It does not have to be an ombudsman as such, in any case. Any kind of independent representative of the public interest will do.

              Your proposed solution (don’t use Google, tell your friends) is welcome, but insufficient, in my view. It is not (only) my privacy I am worried about, or the way I personally receive information. My concern is also about how Google shapes society.

              1. Hi Carissa,

                The two claims you put forward as 1 and 2 don’t really capture my position. It would be better to put it like this:

                1. To the extent that there is a problem with the internet remembering too much (either about content producers (ie right to be forgotten) or about consumers (browsing privacy), it is a sector wide problem, and is most acute for search engines. Any regulation should be focused on the sector, and perhaps the search engines, but not on a specific company.

                2. Any regulation needs to be enforceable, practical (ie low compliance costs), effective and equitable. You haven’t suggested any regulatory package that meets these requirements.

                Let’s consider an actual natural monopoly situation – US railroads in the 19th century. What you’re suggesting is like going heavily after a particular first mover – “let’s get that Vanderbilt!” – rather than trying to regulate the industry as a whole. When regulation did come to the railways, it was reacting to special interest pressure (esp from farmers, who wanted the benefits of the railways without the costs…) Because railways became seen to be of vital public interest, the railways were obliged to do a bunch of things that were costly and which, in the long term, made them (very) uncompetitive with alternative (cars, buses, aircraft). By 1960 the industry was losing so badly much of it was near collapse. Unsurprisingly, deregulation was the answer. This let them concentrate on their core activities (freight) rather than their popular activities with voters (loss-making passenger routes). etc. Basically, I think the model you are proposing sounds a lot like the populist calls for railway regulation in the 19th century. You’re too narrowly focused on existing firms and their market power, and not focused enough on the overall health of the sector. You don’t propose anything generic (like a decay time/auto-delete for webpages) and you’re not focused on consumer protection through any means except the exceptional treatment of one company. I don’t see how that’s ethically defensible, and I don’t see how it could be more broadly effective.

                Finally, I take your point about the role of an Ombudsman in newspaper oversight, but would note the following: (1) Google is a search engine, not a content provider. It’s no more sensible to hold Google accountable for the content of searches than it is to hold the London Underground responsible for the content of the newspapers strewn across it. (2) You haven’t told us what the role of the Ombudsman would actually be within a regulatory package, just that they would “represent the interests of the people”. Who is the people? The global community or a nation’s people, or the people of Europe, or…? And how would the represent those interests?

              2. So, you start off with one regulation, the right to be forgotten. Then you start introducing more “non-dramatic” regulations (eg requiring safety procedures around personal data).
                This is how regulation winds up stifling innovation: it’s like putting stones in a stream. One stone doesn’t block the stream. Enough of them though and you’ve got a dam.

                And what’s your evidence that newspaper regulation works well? Has democracy and fairness gotten better since the newspaper ombudsmen were introduced?

          2. On switching to another provider, that’s what everyone did to switch to Google.

            Google must show an independently appointed ombudsman (or a similar figure) its algorithms. Let’s suppose, for the sake of argument, that Google can be guaranteed absolute confidentiality so that nothing can get leaked to its competitors. Would it really be such a terrible thing? Would it “hobble” Google, “take away” what’s theirs, impede innovation, “kill” the company? I don’t think so.

            So, Google shows the independently-appointed ombudsman Google’s algorithms. The ombudsman says “Man, these suck. Go away and rewrite them to include x, y, and z”. Does Google have to follow the ombudsman’s instructions? If NO, what’s the point of this regulation? To make a nice cushy job for the ombudman? If YES, then yes, it problaby will hobble Google, and impede innovation.

            : Google should not be in charge of requests concerning the right to be forgotten; public officials should deal with that. Would it be so terrible?

            Yeah, it’s not like anything has ever gone wrong with censorship when it’s done by public officials.

      2. Hold on, it was regulation which made Google implement the right to privacy. Google actually went to court repeatedly to fight against that regulation. If this regulation was so good for business, why did Google spend money on lawyers trying to stop it? And, indeed, why are you criticising this regulation in the first place?

        A regulation that puts that burden elsewhere would likely be welcomed by Google. It would certainly save them money, effort, and problems.

        But that money, effort and problems would still be born by someone in society. Meanwhile, if you strike the regulation entirely, all that money, effort and problems goes away.

        they should be made by someone whose priority is the public interest

        And who is this hypothetical person? Working out what the public interest is is incredibly difficult and people often fail at it, which is why we have all the expense and hassle of elections. If there was someone whose priority was the public interest, and we knew them to be it, we could just appoint them as Plato’s philosopher-kings and go home.

        In any case, I stand by my main point: I do not want a private business tailoring what we can see on the Internet through secret algorithms that face no accountability.

        So what are your practical alternatives?

        The situation before Google, when it was far harder to find anything on the internet? The current situation where the government mandates what we can see on the Internet through secret algorithms that face no accountability? The situation where we have much less innovation on the Internet, because we have more government regulation? A situation where we have publicly available algorithms so they can be gamed by spam artists?

  3. Hi, Dave and Tracy. Thanks for your comments.

    The post I wrote was not meant to propose what is the concrete regulation needed. If I felt confident that I knew exactly the regulations that are needed, I would have written a post about that. I do not have all the answers, however. My post was much less ambitious. Its main objective was to make the point that we seem to need regulation, but none of the traditional ways of regulating seem to be appropriate, so we need to think of something new. What that looks like, I’m not sure, but maybe in constructive dialogues such as this one we could come up with some suggestions that might be satisfying. If you are not satisfied with the suggestions I ventured (very tentatively, if I may add), maybe you could venture some other suggestions and we could analyse the merits of those.

    I agree with you, Dave, that regulation should be sector-wise. I picked Google because, while it is technically just another search engine, in practical terms it is more like *the* search engine. I also agree that regulation needs to be enforceable, practical, effective and equitable.

    Citing bad regulation is useful in that it helps us identify what doesn’t work in order to avoid repeating mistakes. Citing bad regulation, however, is no argument against good regulation. Surely some regulation is good, right? Most of us would not prefer to live in countries with no system of laws, we would not want to drive in streets that had no red lights.

    I just doubt we are in the best of possible worlds when it comes to search engines and social media—particularly those that are so big that they can have a huge impact on the world population. I think some regulation that helped protect privacy would make the situation better. I think a public debate about what criteria should be designed into the algorithms for search and news feeds is a good thing.

  4. This commentary is a clear indication as to why nothing innovative comes from the ‘field’ of philosophy. I urge the author to try to stay away from business, lest she be observed as knowing nothing about it.

    1. Hi, Harry the Hat. Actually, a surprising number of people in business majored in philosophy, including some very innovative people such as Peter Thiel. (Not that I am claiming any business knowledge or abilities for myself. I am, however, interested in business and I favour its needs more than perhaps could be glimpsed through this post). Feel free to expand on your comment though—What sorts of things about business should philosophers be aware of? Always happy to learn something new.

Comments are closed.