In this essay, I’ll explain why companies have existed and what characteristics of companies will continue to exist in the coming decades when societies have transitioned towards the post-industrial era. This transformation is also called the Second Machine Age (book) by Andrew McAfee and Erik Brynjolfsson, who I will cite multiple times.
ONE. Are Companies Passé?
Nassim Nicholas Taleb notes in his book Antifragile that the shorter time a specific technology or habit has been around, the more likely it’s to get disrupted and disappear. Chairs are likely to be around for a long time after we no longer need CD players or cars.
There were companies with legal rights conducting business operations in the ancient Rome and ancient India, at the latest around the year 550 AD. So, following the aforementioned rule of thumb by Taleb, companies are quite unlikely to disappear. Nevertheless, Andrew McAfee and Erik Brynjolfsson have dedicated a whole chapter in their new book Machine, Platform, Crowd to discuss the future prospects of companies as a form of future collaboration. The chapter is revealingly named “Are Companies Passé? (hint: no)”.
While making a dichotomy between the core (a company) and the crowd (decentralization), McAfee and Brynjolfsson end up in the outcome revealed already in their headline: companies are still needed in the future. I claim that the answer is not that straightforward: companies might not be passé but companies as we know them shall perish.
The chapter on companies in the Machine, Platform, Crowd starts with a good question: “Do we still need companies?” This essay uses that question as a starting point, broadens and contextualizes the arguments which are mentioned in the Machine, Platform, Crowd and elsewhere, and answers that question in a more thorough and satisfactory manner.
TWO. Why Do Companies Exist? Introducing Two Bad and One Good Reasons
There are three reasons why companies exist, according to Machine, Platform, Crowd.
The first reason is that companies enable long term investments. Because they are “thought to endure indefinitely”, they make it possible to make long term investments and to create long term projects.
This reason isn’t very convincing. Obviously, companies don’t last forever. Moreover, the average lifespan of a company is getting shorter (MIT Tech Review, Innosight [pdf], BCG). Furthermore, in the current quarterly capitalism the focus of companies seems to be getting shorter and shorter. There are many reasons behind this shortening of the corporate time horizon. Short term rewards for CEOs certainly play a role, but it’s also true that companies that focus on short term wins gain more resources for long term struggles — a solid strategy in the ever-changing business environment where paralysis by analysis (Ansoff) threatens those who prepare too much.
So, even if companies exist for long term investments, they do so less and less.
The second reason why companies exist is that companies operate within an existing legal framework and laws. The existing legislation offers more predictability and thus builds on confidence and trust.
I’m not convinced by this argument either. It follows the “It’s always been like this” line of thought which hinders change. While there are significant path dependencies and benefits in the corporate law, there must be a reason outside this legal framework to argue for them. Laws are made by the people: they can be changed. Further, they should be changed if there is a better alternative.
So, even if it’s true that current laws offer predictability and trustworthiness, there needs to be a supporting argument outside the scope of the legal framework itself not to change the laws.
The third reason why companies exist is that they work. They are, as mentioned in the Machine, Platform, Crowd, one of the best ways to get things done in the world. This view can be contradicted for example with points raised by Mariana Mazzucato (pdf). Her argument is that companies do not in fact single-handedly create significant change. Instead, it’s governments who make the change happen, though this happens via companies, through the government investments, military spending and university research.
However, even if we take it as truth that companies simply work, it’s clearly not a satisfactory explanation. We have to ask why companies are one of the best ways to get things done.
“Why companies are one of the best ways to get things done?”
McAfee and Brynjolfsson give one good reason: companies provide residual rights, meaning that they are a way to know, share and act upon clearly defined responsibilities. I will elaborate on this later. On top of the view that companies provide residual rights and are thus needed, I want to add two more arguments for companies: that they currently allow for accumulation of financial capital and that they allow investors to limit the risk they take. I’m going to dive deeper into these characteristics of companies later in this text, but to do so properly, we need to first take a glance at two specific kinds of a company that have nowadays become almost synonymous with the term “company”: The Joint Stock Corporation and Limited Liability Company (LLC).
THREE. History of the Joint Stock Company and Limited Liability Company
The first companies that operated in the way modern mega-corporations operate where the British and Dutch East India Companies. As told by William Dalrymple in the Guardian long form story, British East India Company made an incredible feat: it single-handedly reversed the balance of global trade for several hundred years. It ferried opium to China, and fought the opium wars to seize Hong Kong and claim its monopoly in narcotics, while simultaneously shipping Chinese tea to Boston’s harbour, triggering the American war of independence. Considering India, it wasn’t actually the British who did the conquering, but instead the dangerously unregulated private company with a staff of 35 employees and a small five windows wide office. “Nevertheless, that skeleton staff executed a corporate coup unparalleled in history” Dalrymple writes: “the military conquest, subjugation and plunder of vast tracts of southern Asia.”
East India Company was not the very first joint-stock company, but during the 1600s family partnerships were the typical company structure over most of the globe. As a joint-stock company it was able to issue tradeable shares on the open market. This allowed it to realise much larger amounts of financial capital than the previous legal entities. This financial capital allowed it to grow much larger in impact and scope than the previous companies.
Indeed, when taking a closer look at the British East India Company, one notes that it has remarkably similar characteristics to the mega-corporations of this day. Considering the CEO pay, the most of the looting in India was driven by one man, a sociopath called Robert Clive, who returned to Britain with a huge personal fortune that made him the richest self-made man in Europe. However, his “achievement” wouldn’t have been possible without the reliance on government support that trumps the current Mazzucationist comparisons. East India Company actively used its separation from the government: it managed to secure the involuntary privatisation authored by the Indian Shah as a corporate ownership even though the British government spent vast sums on military operations to protect the corporation’s acquisitions in India.
East India Company was also almost ridiculously short-sighted in its operations. Very much like the current day mega-corporations, it proved very vulnerable to economic uncertainty. Dalrymple writes:
“Only seven years after the granting of the Diwani, when the company’s share price had doubled overnight after it acquired the wealth of the treasury of Bengal, the East India bubble burst after plunder and famine in Bengal led to massive shortfalls in expected land revenues. The [East India Company] was left with debts of £1.5m and a bill of £1m unpaid tax owed to the Crown. When knowledge of this became public, 30 banks collapsed like dominoes across Europe, bringing trade to a standstill. — The East India Company really was too big to fail. So, it was that in 1773, the world’s first aggressive multinational corporation was saved by history’s first mega-bailout — the first example of a nation state extracting, as its price for saving a failing corporation, the right to regulate and severely rein it in.”
In a sense, this first bailout was one of the first times that law granted limited liability, albeit posthumously, to a company, because the company ownership was not, to the full extent, liable for the atrocities and economic turmoil that it caused. Nevertheless, it took 200 more years before ex ante limited liability became the standard legal procedure in companies. First incorporated in law in New York in the year 1811, it became the norm by the end of the century. In a way, limited liability is the joint-stock corporation on steroids: on top of providing a means to collect vast sums of financial capital, which is allowed already by joint-stock legislation, it also offers a legal protection against risk to these investments. In a limited liability company, the stock owner is only responsible for the invested amount, even if the company conducts atrocities that lead to claims larger than its financial capital. Thus, limited liability makes corporations legal entities in the similar manner as people, and grants the owners a risk imbalance which only a few centuries ago would have been considered both counterintuitive and plainly unfair.
The move towards limited liability companies was thus not accepted without quarrel. There was an argument that it would lead to a drop in the standards of probity. Nevertheless, the benefits of this new piece of legislation were so remarkable that the counterarguments were pushed aside. Without the risk of being personally responsible for the mistakes made by the companies, the amount of the available financial capital vastly increased. Paired with novel technological opportunities, this new legal form with mitigated risks and piles of financial capital pushed the companies towards large-scale industrial enterprises that we are familiar with today.
In retrospect, the limited liability company became the modus operandi of the industrial era. It allowed for building large factories to support vague but convincing ideas, created venture capital structures, shaped universities to focus on educating for work rather than for research, and allowed for accumulation of corporate profits to ever smaller number of benefactors via aiding big creditors instead of the small ones in the unfortunate but typical event of a bankruptcy.
Thus, the limited liability company was, without a doubt, one of the key driving forces behind industrialisation and the new global world order where floating fiat currencies, growing economic inequalities, global capital, neo-colonialism, technological planetarism, and capitalist, and mass media controlled democracy dominate the life of a common man. It allowed for great prosperity and wellbeing by facilitating new services and products in the market. On the other hand, the limited liability company enhanced economic inequalities and let investors avoid responsibilities when a company they invested in conducted wrongdoings.
In conclusion, unlike the history of the company which goes back thousands of years, the limited liability company is a fairly recent legal institution. It has become in essence naturalised to a default for of doing business, by which I mean that the limited liability company feels like a natural part of the composition by which the institutions operate. It’s hard to see through this naturalisation and understand the limited liability company in its historical context. This naturalisation can cause myopia when trying to grasp the characteristics of companies in the post-industrial era.
FOUR. Arguments for the Limited Liability Company
As I mentioned previously, there are three arguments for the limited liability company.
First, the limited liability company is a very efficient method for accumulating financial capital. This has been crucial for companies, because in order to create products, they need to invent, produce and advertise them. To produce goods, one needs either to make purchasing contracts or even to build a factory. And to advertise a product on the market, one is required to reach potential buyers and convince them of the benefits of the product — an act which is uncertain to succeed to say the least. Furthermore, an entrepreneur might need to also create and develop a new product. Innovation, marketing and production are all very costly and risky acts. Nevertheless, because there is often financial capital available, a convincing entrepreneur with a sufficiently easy or ridiculously brave idea can usually get the necessary funds to try it out.
The second argument for the limited liability company is that it offers legal protection against risk. Or, to be more precise, limited liability puts a ceiling on the risk, as the investor can only lose the invested amount, not more. This aspect feels natural these days, but it’s in fact the most unnatural aspect of the legislation: there is no one responsible for the totality of risks that result from the operations of a company. In the early days of the limited liability company there was a habit in Britain that owners would buy limited liability only for a part of their shares. The rest of the shares came with full liability to risk.
Third, the limited liability company provides a clear set of rules on how to divide residual rights and work between the owners, other stakeholders, and the management.
Next, let’s take a look at these arguments one by one by integrating them into a set of recent developments in the current operating environment of the limited liability company. This allows us to understand the potential disparagements between the industrial company, which is in essence the joint-stock limited liability company, and the post-industrial company.
FIVE. Capital Accumulation in the Age of the Superabundant Capital
As Manking, Harris and Harding write in the Harvard Business Review, in the past, business leaders have viewed financial capital as their most precious resource, but today, while accumulation of financial capital continues to be important for companies, it’s no more the most important asset companies have. Financial capital is abundant and cheap: global financial capital has more than tripled in the last 30 years and its price has plummeted: real borrowing costs are today very close to zero. This means that any profitable enterprise can get the capital it needs for product development, production and market entry.
Manking, Harris and Harding claim that the era of superabundant capital will continue for at least 20 years. This is mostly due to the fact that globally the number of people around their fifties will increase. The outcome is that superabundant, cheap capital is not likely to disappear any time soon.
This shift is immense. For existing companies, it completely changes the strategy playbook. Nevertheless, the effect of the superabundant capital to longer term change is even more thorough, because it undermines one of the basic arguments behind the limited liability company, and further, even the joint-stock company.
According to Manking, Harris and Harding, the assets that are in short supply, rather than financial capital, are the skills and capabilities required to translate ideas to products and services. We’ll take a look at this point of view in the final chapter.
SIX. Mitigation of Uncertainty in the Age of Big Data and Machine Learning
As mentioned above, limited liability means in practice that the investor knows a hard ceiling for the risk she is carrying with a particular investment: she can never lose more than she invested.
To get hold of how the formulation of risk is changing, we need to make a distinction between two kinds of categories for unknown futures that are often confused: first, there is uncertainty, and second, there is risk.
When we know what the distribution of possible events looks like, we are talking about risk. Risk in stock is the same as variance: high risk means potentially high reward.
When we don’t know what the distribution of possible events looks like, we are talking about uncertainty. This is the case when we have no idea of what the possible outcome of an event might be.
Advanced analytics including machine learning, combined with a vast and growing set of different data sources, including real world Internet of things sensors, constantly increase transparency. This means that more and more topics and themes move from the category “uncertain” to the category “risk”. Further, we might also get better at understanding different risk distributions, but this is very much a secondary goal.
The real purpose of the limited liability legislation is not to get rid of the risks in business, because, as I mentioned, risk is the same as variance: the higher the risk, the higher the potential reward. Rather, the goal with limited liability company legislation has always been to limit the uncertainties which are built in all the real-world operations that businesses inexorably are. The increase in data analytics, sensors and data reduce the need for this kind of uncertainty management.
When the probability distribution is known, the investors can calculate the risk they are willing to take even if it’s bigger than their initial investment in the company. This is especially the case in the modern day investing where practically all portfolios from the smallest individual investors to the largest hedge funds aim to diverge their investments. If the probability distribution is taken care of and the distribution of investments is large enough, there is no need to limit the ceiling for risk. Nonetheless, there is always some amount of uncertainty in any investment, and this should be taken care of. Clearly, however, the inbuilt uncertainty in investments is getting smaller, which reduces the need for legal structures such as the limited liability company.
To conclude, the assets required by the company to be as tempting as possible for a potential investor is less and less a tempting narrative or value proposition, and more and more often better data and analytics capabilities to move as much uncertainty as possible within the known risk distribution. We’ll take a closer look at what this means in the final chapter.
SEVEN. Residual rights
In Machine, Platform, Crowd McAfee and Brynjolfsson return to the classic question by Coase (pdf) on the nature of the firm. The question by Coase was “if markets are great, why are there companies?” The authors elaborate by writing that “why, in other words, do we choose to conduct so much economic activity within these stable, hierarchical, often large and bureaucratic structures called companies, rather than just work as independent freelancers, coming together as needed and for only as long as necessary to complete a particular project, then going our own way afterwards.”
The answer to this question by Coase was that markets do have higher costs than companies in many areas, such as in searching and discovering relevant prices, as well as in negotiating and making decisions and in making and enforcing contracts.
Clearly, McAfee and Brynjolfsson point out that, while paraphrasing a typical technologist, digital technologies reduce costs in many of these segments. They cite examples from the failed DAO experiment (link) and conclude that even if it failed, there might be a better one coming shortly. They then turn to transaction cost economics, which seems to contradict their earlier argument by showing that companies are still needed.
Why? According to Transaction Cost Economics and especially citing the Finnish Nobel prize winner Bengt Holmström, McAfee and Brynjolfsson point out that companies are structures that grant residual rights, meaning that they provide rights for owners regarding things that are not specified in contracts. The argument made in Machine, Platform, Crowd is that complete contracts, such as DAO, are not possible, so there is a role for the company in the future as well. I won’t go deeper into the argument on why complete contracts are not possible, as it seems rather obvious to me that the inherent uncertainty, albeit reduced, will always cause previously unseen circumstances to arise.
Similarly to the last point, the authors of Machine, Platform, Crowd seem to indicate that new innovations in IoT and computing power make more complete contracts possible and lower production costs, which would make markets more lucrative in comparison to companies and other hierarchies.
However, there’s a point which is not mentioned in the book: while digital technologies reduce the costs of markets by lowering the production costs, these technologies also simultaneously reduce coordination costs in hierarchies. For example, a large company can soon install a voice recognition and keyword analysis system to all of its meeting rooms to algorithmically make sure that people who that talk about similar things in their respective corporate silos know about each others.
So, it’s difficult to find out if market gains ground in comparison to companies purely based on the technological transition. Indeed, supporting this view, McAfee and Brynjolfsson point out that recent economic data does not show that the number or the size of companies would be diminishing.
Further, they point to anecdotal evidence, writing
“[Platform] companies are trying to make it as easy as possible for some types of partners to enter and leave a business relationship with them, giving rise to the notion of an ‘on-demand economy.’ Other companies are exploring how to deliver value with blockchains, smart contracts, and other technologies of extreme decentralization. But they’re almost all pursuing these radical goals within the highly traditional structure of a joint-stock company, an organizational form that has existed for well over four centuries. When we visit these companies, we’re struck by how normal they look. They all have employees, job titles, managers and executives. — “
McAfee and Brynjolfsson end their chapter on the future of the company in a rather vague conclusion: “Leading companies of the second machine age may look very different from these of the industrial era, but they will almost all be easily recognisable as companies”. However, we cannot leave the investigation of the post-industrial company to such an empty statement! Clearly, either the companies look very different or they don’t. So, which one is it? Which characteristics of the current companies will remain and which ones will not?
Returning to the anecdote by Nassim Nicholas Taleb mentioned in the beginning of this essay, the longer something has been around, the less likely it is to disappear. Residual rights are a feature of all companies and thus they are not limited to just limited liability companies or joint-stock corporations. Thus, my answer to the question “Do we still need companies?” is “Yes. We have had companies for thousands of years and there will be companies in the post-industrial era, because there needs to be a way to accumulate various resources and provide residual rights. But we also need to get rid of the industrial era legislation that hinders the ethical, collaborative and efficiency capabilities of the post-industrial companies. With this I refer particularly to the limited liability company.” If this is indeed the case, what could be the characteristics of the post-industrial company?
We need to get rid of the industrial era legislation that hinders the ethical, collaborative and efficiency capabilities of the post-industrial companies.
To conclude this chapter, I claim that as there is no way to make complete contracts, there will be a need to provide residual rights in some way. Thus, if we decide to define companies as organisations that provide residual rights, there will be companies in the post-industrial era. However, I disagree with the view that they will “pursue these radical goals within the highly traditional structure of a joint-stock company”. There is a need and opportunities for new ways to divide residual rights and to allocate people and assets in companies that no longer prioritise financial capital and that mitigate uncertainty with ever-improving data and analytics. I shall return to this point of view in the next, final, chapter.
EIGHT. The Post-industrial Company
The limited liability company was born rather randomly to circumvent two crucial bottlenecks that companies faced in the dawn of the industrial era: the accumulation of capital and the mitigation of risk. We don’t yet know what the bottlenecks of the profit models of the post-industrial era companies are but I have presented a few alternatives in this text. Further, we don’t know what the winning business system in the post-industrial era will be, but there are a few candidates already: platforms (read my text on platform business models), co-ops and collaboration networks.
The required assets for the post-industrial era profit models that I have argued for in this text are data and analytics capabilities to hinder the risks, skills and dynamic capabilities to enable efficient utilisation of abundant capital, and new modes of leadership and management that I speculate would focus more on roles than people and more on culture than practices (read my text on leading a dynamically changing team roles with culture interventions) to divide residual rights and accumulate resources more efficiently. This is the shape of the post-industrial company.
With these assets, post-industrial companies can operate without the shield of limited liability, which, as pronounced, has had great merits but also severe pitfalls, especially because it structurally diminishes advocacy. In the lack of global regulative action, limited liability companies are likely to remain as the default form of company for several decades. However, the time when the limited liability company was the clearly superior mode of organising is over.
To conclude, companies are not passé. They are one of the best ways to change the world and do good things, because they are a good way to divide residual rights and to coordinate some operations more efficiently than the market. However, the modus operandi of the industrial era company, the limited liability company, has several pitfalls. These pitfalls include increasing financial inequalities in the form of benefiting large creditors, allowing investors to have a ceiling for uncertainty and making no one fully responsible for the bad acts of the companies. Previously, limited liability companies have existed for three reasons: to accumulate capital, to mitigate uncertainty, and to share residual rights. Nevertheless, the capital accumulation is no longer a bottleneck in companies, and there’s less uncertainty to mitigate because data, sensors and analytics turn unknown uncertainties into known risks where the risk distribution is understood. Regarding residual rights and resource accumulation, there might be more efficient ways to operate and share them than an industrial era hierarchical corporation. Current candidates for a post-industrial era business system include platforms, networks and co-ops. Further, there needs to be a model for the post-industrial era to tackle the most important bottlenecks. The current candidates for the bottlenecks are data, analytics, skills, dynamic capabilities and new modes of leadership and management, including platform management.
Clearly, companies are not passé, but they won’t be what they used to be, either.
Discussions with Juha Leppänen contributed significantly to this text.