business firms

/ / , , , , /

Complexity, Institutions and Firms

Are they associated?

We all know that one of the central problems of economics is the existence of uncertainty.  At least since Frank Knight’s work in the 1920s, uncertainty has been something of concern to economists.  Knight’s description of uncertainty as being a condition in which no probability distribution existed, or could exist, has led some of the most eminent theorists of economics to argue that theorizing is simply not possible.  Which is a rather formidable obstacle.  I assume they meant that their particular form of theorizing was not possible.

One person who rose to the challenge represented by uncertainty was Douglass North, who famously built his theory of institutional change to show how humans respond to  the endemic uncertainty they deal with on a day-to-day basis.

I am going to use three separate quotes from North’s book “Understanding the Process of Economic Change” as a starting point for this short discussion.

First:

“But uncertainty is not an unusual condition; it has been the underlying condition responsible for the evolving structure of human organization throughout history and pre-history”

Exactly.  But the mere existence of uncertainty is insufficient to explain modern economic institutions — or any other institutions for that matter.  There has to be more to it.  He goes on to reference Ronald Heiner who identified that the gap between a person’s competence and the difficulty of the problem being solved.  Here is North picking up the story:

“The human agent in the face of such a gap will construct rules to restrict the flexibility of the choices in such situations.  We know these rules as institutions.  By channelling choices into a smaller set of actions, institutions can improve the ability of the agent to control the environment.”

Very true.  Anyone who has lived in the business world will recognize themselves in this situation.  The whole point of containing the entire process of production within an organizational boundary is to set limits on the variability of outcomes.  Which is something the vastness of a market cannot do — at least outside the imaginations of economists.  North describes it as controlling the environment.  It is this need to reduce the options available to just a few specified by the plan the business is trying to execute that is the essence for the existence of a firm.  In the surrounding swirl of uncertainty it would be impossible to construct the products and services that businesses routinely produce.

But one person’s control adds to another person’s uncertainty.  North goes on:

“While the deep underlying source of institutions has been and continues to be the effort by humans to structure the environment to make it more predictable, this effort can and frequently does make for increased uncertainty for some of the players.”

This is intriguing.  It is also familiar to anyone in business.  Not only do I as a manager have to take into account and limit the problems presented by uncertainty in my surroundings, I have to take into account the actions of my competitors who are also trying to control uncertainty.  Life gets complicated.  We are all caught up in the swirl.  We make it worse for others as we each try to tame it for ourselves.  It’s as if the amount of uncertainty is a constant and that by controlling it locally we simply export it back into the wider environment.  Maybe there’s a law of economics somewhere in this that needs articulating.

But this raises an interesting question: what is the relationship between uncertainty and complexity?  Are they relates?  Are they the same thing?  Are they just two concepts passing each other in the night?

North gives us a clue — the rest of his book is his explanation.  But let’s move onto our own speculation:

By creating structures to contain and thus localize our environments we multiply the loci of uncertainty.  We restrict it within the boundary of the structure, but we add to it outside the structure.  This sets in motion a competition to create more structure and thus to add to the multiplication of uncertainty.

But this simple process is simultaneously adding to the complexity of the known environment.

Each structure is created to carry sufficient useful information to accomplish whatever task the structures creators intended.  In the case of a business firm the production of something.    Setting the boundary is one of the most important tasks a manager has to accomplish.  What ought to sit inside the boundary?  What can be left outside?  Structures like business firms hoard information in order to maintain their own identity and ability to differentiate themselves.  They become repositories of information.

Of course the information constantly leaks out into the surrounding environment: whenever a product is sold it carries along with it its design so someone on the outside can reverse engineer it and extract that information.  The quantity of information in our environment is constantly rising.  This is one way it does.

This is, however not our central question.  Our real concern here is, what explains the emergence of institutions through time?  Why, if North is correct, did things like our current business firms not exist in prior ages?   Was uncertainty less of a problem?  Was there less uncertainty?

 

No.  I don’t think so.  And this is my quick way of associating complexity with uncertainty, at least for our purposes:  as we expand the circle of our knowledge we simultaneously make the management of what we know more complex.  We increase the number of inter-relationships.  We make our decision making more difficult simply by revealing more to ourselves and increasing the number of options available.  We have great deal more information to process.  This is, of course, just a modification of the ages old paradox that the more we know the less we know.  Except in this case we are trying to make the paradox applicable to our explanations of economic growth and the development of structures within the economy.

North focuses on uncertainty as the underlying cause for the creation of institutions.  But that leaves us only half way to understanding why things like business firms exist.  Yes, they deal with uncertainty.  But uncertainty is endemic.  It is fact of life.  Why did firms become more important as the economy grew beyond its agricultural and early industrial roots?  Because the rising complexity created by each wave of structures necessitated a new layer of structure to handle the flood of information.

Uncertainty is in this sense a constant.  Complexity is not.  As complexity rises we should see an institutional response.  The history of organization and management suggest this is true.  And the current transition into a more digital economy suggests that we will need newer structures in the future.  Will, for instance, older business firms be replaced by networks?

Or will older firms find ways to simplify things sufficiently to remain relevant?  It is, after all, the role of a firm to simplify and codify its information so it can replicate its products easily.

Which is where machines enter the story.

/ / , , , , , , , /

1937

Hayek says this:

“The problem which we pretend [to] solve is how the spontaneous interaction of a number of people, each possessing only bits of knowledge, brings about a state of affairs in which prices correspond to costs, etc., and which could be brought about by deliberate direction only by somebody who possesses the combined knowledge of all those individuals … “

This is from his essay ‘Economics and Knowledge’ which was published in 1937.

Hayek’s thrust is, of course, to demolish the notion that an economy can be centrally planned to any degree of efficiency. His argument is simplicity itself: the body of knowledge that exists within an economy is so asymmetrically distributed, so extensive, and so diverse, that it is not possible for anyone sitting at the center to know whether their proposals and policies are the “best” solution to whatever problem is being solved.

The “we” in that first sentence is the economics profession — his essay is based on his 1936 presidential address to the London Economic Club, so he addressing an audience of economists and attacking their assumption that everyone knows everything, which is really an assumption to make modeling more tractable rather than an effort to understand the world.

And I think he’s correct.

Knowledge is scattered about the economic landscape. It is clumpy rather than smoothly available to everyone. People hoard and protect knowledge in order to extract value from it. A pictorial representation of the spread of knowledge would look decidedly granular.

But having made this point, Hayek then goes on to commit an error:

” … And experience shows us that something of this sort does happen, since the empirical observation that prices do tend to correspond to costs was the beginning of the science.”

No, it doesn’t.

Having established the impossibility of knowing everything, Hayek then slips back and appears to assert that we do know something: he is telling us that prices correspond to costs.

But, if certain knowledge is always going to remain inscrutable to us, how would we know that? We cannot. Hayek is contradicting his own argument.

And this, in essence, remains an issue in economics. The asymmetrical distribution of knowledge, which allows the creation of value to those who use their knowledge for that effect, is the driver of economic activity. Economists, in order to build their elegant and rigorous models, have eliminated this asymmetry. Why? Because the clumpiness of knowledge is the telltale sign of fundamental uncertainty, and fundamental uncertainty is a difficult subject to deal with if you want to build models that automatically tend towards equilibrium. In the face of uncertainty you can never tell where the tendency is headed. Or why it is headed there.

Later on, in the same essay, Hayek says this:

“Clearly there is here a problem of the division of knowledge which is quite analogous to, and at least as important as, the problem of the division of labour. But while the latter has been one of the main subjects of investigation ever since the beginning of our science, the former has been as completely neglected, although it seems to me to be the really central problem of economics as a social science.”

Is he correct in this?

I am not so sure. I see the two as being the same thing. To me, the division of labor is an instance of the division of knowledge. Specialization is a consequence of, and a contributor to, learning or skill development. The more components there are, the more opportunity there is to seek value from each.

In any case, this division of knowledge, and the profound asymmetry it speaks to along with the accompanying uncertainty it engenders, is why business firms exist. Firms collect knowledge, they protect it, and they deploy it for profit. They actively seek to make the economic landscape more clumpy rather than the perfectly smooth plane that economists love to postulate.

Which is why, in my opinion, to understand the economy we must understand the business firm. This is not an advocacy of the “micro-foundations” school of thought, it is simply a recognition that business firms prowl around the economic landscape, doing their business — for better or worse — and we thus need to accommodate them centrally in our thinking.

Ironically, also in 1937, another economist was making this point. This was the year that Coase launched the transaction cost view of the firm. His central question was, in short: if markets are as perfect as economists make them out to be, why are some economic activities undertaken within the boundaries of firms? This ought not to happen: there should be no gain from enclosing some activities within firm boundaries. So what’s going on?

Uncertainty, and what Coase calls the “institutional structure of production”.

Production needs structuring within institutions precisely because fundamental uncertainty precludes alternatives. This is especially so in more intensely clumpy knowledge landscapes. Getting control over the entire production process is a method that mitigates the impact of uncertainty. At least you can get the product made reliably, both in terms of consistency of quality and volume. It can be no accident that, as knowledge became more specialized and thus more valuable to concentrate, and as products became more complex through time and space, more control was needed.

The outcome is what we can describe as “localized central planning”.

Which is what Hayek would have concluded too, had he not been hellbent on attacking universal central planning.

In any case I find it fascinating that the two, Hayek and Coase, both in their own way, brought the impact of uncertainty to the fore in the same year.

It’s a shame that economics has never fully embraced, nor realized, the full richness of their ideas. Neither author was willing to step into the world that they clearly understood existed. Hayek was right about universal central planning: it is an impossibility. He was wrong to assert that this implied anything about the market place or prices. By his own argument we simply cannot know whether something is optimal. Uncertainty makes such a thing inscrutable too us. And Coase was equally correct when he saw the need for local central planning: it is the only way we can organize production adequately in the face of uncertainty. But his focus on transactions was a legacy of the classical emphasis on exchange. It ignored the need for active coordination. He missed the requirement for management. He should have talked about “management cost” not “transaction cost”. They’re different animals.

So: an interesting question is this: what happens to Coase’s “institutional structure of production” when information, and by association knowledge, is less clumpy in the economic landscape? Does something like the Internet, which is a vector for information and knowledge, obviate the need for such structure? Does it smooth that landscape out sufficiently for firms not to exist?

We need to think about that.

We need a new version of the discussion that ought to have taken place in 1937.

 

/ / , , , , /

Operating Space – First Cut

 

It has become impossible, apparently, for anyone to do a sober analysis of the torrent of fads that collectively are known as the start-ups of Silicon Valley. Instead we are given a fast paced – it has to be fast paced nowadays for fear of being disrupted five minutes later – and inevitably glowing surface-skim of the impact all these fads will have on the economy.

All we know is that the change will be total. Absolutely total.

Right.

It seems like an age ago, but there was a time when I was deeply interested in the impact that the internet and related technologies would have on business. I thought about the problem for a long time. Indeed it was the reason I re-connected with economics because I, rather foolishly in retrospect, assumed that economics would be a good source of knowledge about the fundamentals of business.

I began with the question: “why do firms exist?”.

And there I ended.

I ran into what I now call the ‘Coase Conundrum’, which is simply that economists go on endlessly about the perfection of markets and then sidestep the rather awkward issue of the presence of huge business firms lumbering about the economic landscape doing whatever they can to despoil all that perfection. The firms succeed. They proliferate. They trample the tender attributes of markets so thoroughly that the economist’s analysis looks not only quaint, but absurd when set against reality.

As Coase so neatly put it: if markets are what economists say they are, then firms cannot exist – at least not as they do. The economists have, naturally, persisted in getting around this conundrum by ignoring it and setting the extensive task of production within a stone age relic called a ‘production function’. This absurdity serves its purpose and insulates economics from  having to deal with the great economic process of creativity. After all production is a creative process. Literally. We create things from other things.

A production function vastly simplifies creativity. It eliminates any role for brainpower, thinking, innovation, or any other category of the application of knowledge, and leaps straight to the raw materials. It posits inputs as being labor, land, and capital.

Think about those three things. Labor, land, and capital. How simple. And how utterly misleading. Not one of them can be distilled into a single attribute. They are all immensely diverse categories of other stuff. Is labor really just brute force or muscle power? In which case it is better thought of as energy doing work. Is land another word for raw materials like coal or iron? In which case why not specify raw materials? Land, after all has long been associated with power and hierarchy, so why isn’t power a component of production? We all know that control is an essential ingredient of the capitalist economy, so why not admit it? And capital? Enough said: the great controversy over capital and its diversity serves us well as an indicator of the true devotion to rigor in economics. There isn’t much. Because capital is such a broad and diverse category the very notion of a single ‘capital’ has no meaning – unless we specify exactly which sub-category we are talking about – but that’s too hard to do in our usual equations. So economics ignores the problem and plods ahead with its rigor-less singularity.

One consequence of this mess is that economics has a frightful time accounting for growth. After all growth is a dynamic outcome of creativity. Since production functions are so littered with ambiguity and poor definition of the real inputs – raw materials, energy to do work, and the knowledge to extract products from a combination of work and raw materials – economics cannot track economic growth very well at all. It has to fudge. Which it does well. When it comes to the production side of an economy economics is nothing but a big brilliant fudge.

Which gets us back to breathlessness.

Because economics cannot account for the existence or function of business firms, it cannot account for the kind of change that seems to be underway as a result of the rise of digital rather than analog processes.

My original stab at an explanation went something like this:

Firms exist to concentrate and exploit information. They thus try to inhibit the perfection of the market. They do this by using their power to assert rights over information: copyrights and patents are viciously defended from encroachment and the laws are changed periodically – under pressure from business – to prolong the ability to extract excess profit from the defense of such information. Put more bluntly: business are in competition with their customers for the information used in production. Customers would ‘prefer’ to use that information directly and thus dispense with the obvious costs of having a business from act as an intermediary.

Set that aside for a moment.

Information is used to create product. Information is found everywhere in business: in the design of products, in the design of production processes, in the design of plant and equipment, in the know-how of employees and managers, and so on. Information is the lifeblood of business. It is safe to say that without information business is hard if not impossible to conduct.

Which is why businesses go to great lengths to store, protect, and develop it. Businesses can be thought of as giant stores which stand out on a landscape of much less concentrated information.

So, far from the immense landscape of flat and equally available information posited by economics, our actual landscape is punctuated and characterized by a great lumpiness. The neat and easily theorized symmetry upon which economics rests does not exist. Indeed, I would argue that it was the breaking of that symmetry and the consequent accumulation of asymmetries of various kinds that allowed the rise of our modern economies.

So what about this breathlessness?

If information is the key to business, then its movement from an analog expression to a digital expression must have implications for the way in which business can be – or will be – conducted. In a digital world the storage of information becomes much more simple. So does the transport and analysis of that information. So, therefore, does the extraction of value from creation. The pace of innovation thus ought to accelerate. Which it has. We live in an age of creativity. Indeed so much is being created that we now realize that much of what can be created is rather banal.

But, there’s a downside to this.

Firms are centers of information. They are hoards of information. In a digital world it becomes much more difficult to protect that hoard. It dissipates more easily. It is more tractable to emulation and thus propagation. Information, as we know, is not used up during production. It persists and can be copied. This allows more easy competition and replication. This implies that the life span of businesses ought to shorten. Which is what, I think we are beginning to detect in the reality of places like Silicon Valley.

A more digital world is thus, inevitably, a world of much less stability and of greater operating  risk.

Which gets us back to that competition between business and its customers. One of my ways to describe all this was to posit the existence of what I call ‘operating space’.

In older economies the space occupied by resources and resources owners was largely coterminous with that of the space occupied by consumption: there was a far greater self-sufficiency. As the division of labor fragmented this older economy and subsequent the pace of creative change accelerated we needed a space in which to to produce. This intermediate space enclosed all the emerging production processes or operations of business in what was becoming a much more come world. Hence my term ‘operating space’.

So business firms occupy this operating space which separates the two older resource and consumption spaces. Indeed it is a burden upon them since it adds cost to manage and protect that operating space and thus diminishes the value each of the others can claim for itself.

Because the activities within operating space are adding value via creativity and the production of novelty unaccessible in the two older spaces, the loss of value by the cost of creating operating space was more than offset by its value added.

Thus operating space was tolerable: as long as it added more than it cost society would benefit.

Th emergence of digital forms of storing and using information have begun to erode this balance. Operating space needs to become much smaller in order to continue to add benefit. The rise of readily accessed, deployed, and used, information is undermining the logic of operating space. Business needs to be compressed to take up less space. It has to be leaner, less intrusive, and more productive. It cannot rely on its older ability to hoard information. Not, at least, as well as in the past. So it has to rely on greater creativity to justify its existence.

One way to reduce operating space is to disassemble the logical sequence of operations needed to in production. Some of those operations can be outsourced into the consumption space. In other words businesses can involve customers on the creation of the product or service. We have seen a rapid growth of this kind of activity: the automation of many service tasks is an example (self service check-out; ATM machines and so on). Theses a way business can offload expensive aspects of operating space onto customers whilst maintaining possession of the core of value creation.

Now this core is under attack also.

An obvious example would be three-D printing, where a customer with appropriate access to information and resources can produce an ever growing variety of products previously attainable only via traditional business processes.

So to recapitulate: in older more self-sufficient economies the need for a large and expensive operating space was minimal. It became both more necessary and more easily justified as the division of labor and information proliferated. Modern business grew as a method for storing, protecting, and using this information. Businesses thus sought to defeat the perfection of the market – and continue to do so, which is why textbook economics is an inaccurate portrayal of real economies.

But the transfer of analog methods to their digital counterparts is eroding the size of the allowable operating space an economy will tolerate. This is a threat to the organization of business as we know it.

Does it imply that the perfect markets posited by economics will be revealed as operating space shrinks under the cost pressure released by digitization?

That we can talk about in the future.

/ / , , /

The Firm – A Note To Myself, Part One

I am thinking through my views on the business firm as an economic entity. My notes so far [no particular order]:

  1. The modern business firm is a reaction to, and exist because of, uncertainty. Other rationales are subsidiary to this. For instance New Institutional Economics [NIE] explains firms in the tradition of Ronald Coase and focuses on transaction costs, agency issues, the potential for cheating, and asset specificity. But all these are symptoms of uncertainty. The problem with NIE is that it tries to hew close to the neoclassical line and thus avoids confronting uncertainty directly. Those symptoms are all useful to know and have to be dealt with, but are not the cause of a firm’s existence.
  2. All firms have a great deal in common – structure, operations, purpose, attitude towards customers, relations with the environment etc – so can speak of each firm as being an “instance” of a general form. We are interested in that general form.
  3. All firms enclose and occupy a logical and physical space in order to offset uncertainty. I call this space “operating space” [OS].
  4. Within OS firms create an “alternative reality” that is more certain than the exterior real world. They do this by invoking strategic plans that substitute arbitrary or created “facts” for the unknowns in the environment. They then assume the reality of these facts and conduct their operations accordingly. The contradiction of the plans by unfolding actual reality puts a premium on adaptability and causes a constant need to update plans.
  5. As an aspect of that alternative reality, firms set prices – in a variety of arbitrary ways – in order to impose some sense of certainty on their immediate future. Such prices are modified as the firm learns about unfolding reality. This sets in motion an iterative feedback loop tying the firm’s planning to the environment – which includes other firms. So firms try to set prices, which end being modified as information is revealed. The dominant force in pricing is their initial determination which is institutional in character.
  6. OS is extensive and thus expensive. This imposes a cost on consumers who tolerate OS provided that it is socially beneficial.
  7. The social benefit of OS is that it allows productive processes that are fragile with respect to uncertainty to take place. The management cost of OS is therefore acceptable depending on that fragility, which in turn depends upon the complexity of the product and the time it takes to produce.
  8. Management cost is not that associated with production, but is – more importantly – that associated with creating, monitoring, adjusting, and executing plans. Management not only manages production, but in conforms that production to the firm’s planned reality.
  9. Any phenomenon that reduces endemic environmental uncertainty diminishes the need for firms – the Internet is an example.
  10. This is because consumers have access to more information and thus causing the cost of creating and maintaining OS to exceed its social value. Consumers then have the ability to “dis-intermediate” OS. This puts pressure on management to eliminate costs.
  11. Enclosed within OS are a set of “processes” that encompass production.
  12. Each process is composed of a mix of “roles” and “routines” that embody the firm’s knowledge.
  13. Roles, consist in main, of “secondary knowledge”, which is the capacity to learn, adapt, and modify actions. It is responsive to the environment, to error, to mutation, and thus is the center of the ability to change.
  14. Routines consist, in the main, of “primary knowledge”, which is the codified, and thus less expensive and more easily replicated, form of knowledge. Routines are specific to a task and require less skill to execute. They are more easily automated. Firms prefer primary knowledge because it is predictable and can thus be built into a plan more comfortably – the output is also predictable, reducing uncertainty with respect to product.
  15. Routines are preferred because of this reduction in uncertainty, but are flawed – perhaps fatally so – with respect to a sharp contradiction not plans by unfolding reality. Primary knowledge is useful only insofar as the future mimics the past, which given endemic uncertainty degrades its use inevitably. Thus the need for the learning capacity of secondary knowledge and its ability to identify the need for and subsequently develop new primary knowledge.
  16. Production itself consist of the imposition of order on hitherto disordered raw material substrate. Thus production is a creative force temporarily offsetting the tendency towards disorder in the environment. Order is imposed by the elimination of information from the substrate. Of all the potential states for that substrate only one is chosen – the others are removed or prohibited. The raw material is thus rendered into a specific state for a single or limited range of uses.
  17. The imposition of order requires energy, which is thus part of the production “function” along with raw material, the knowledge of the required order, the knowledge of how to impose order, and the labor and equipment needed to execute the roles and routines embodying the process.
  18. The production process can be viewed as a lengthy algorithm being applied against the substrate. The longer the algorithm the more information content the end product possesses.
  19. There is a tendency towards greater information content as the accretion of knowledge within an economy enables exploration of more intractable areas of “product space”.
  20. Potential product space has no bounds. Possible product pace is bounded by the constraints of the economy’s “basis”. Actual product space is that part of the possible subject to the plans established within the total set of firms’ OS.
  21. An economy’s basis consists of its state of knowledge [all forms]; its institutional structures; its geographic [resource] limits; its available technology; its cultural constitution; and its available networks through which communication [access to information, contacts etc] and logistics [transport etc] are enabled.
  22. The more complex the basis the more able an economy is to access product space. In particular as economy’s learn there is an increase in complexity in the economy. Rising complexity places a greater premium on OS within which to manage creative processes. Hence the increasing dominance of large OS spaces and rising management cost in general. For society to benefit from the production of complex products it has to tolerate this rise in management cost.
  23. But the emphasis is on tolerance. Consumers would prefer lower costs and smaller OS.
  24. Realizing this, and yet faced with increasing complexity, managements execute a variety of approaches to reduce overall cost. There are two generic approaches: greater emphasis on primary knowledge – lower production cost, but less adaptability; and reducing management cost itself by automating semi-skilled or even skilled activities [converting roles into routines], de-layering, and relying less on internally generated “know-how” and more on externally provided “know-how”.
  25. From this we can understand that business firms – because of their desire to maintain an OS – are in confrontation or competition with consumers. For consumers firms are both desirable [production of complex products] and undesirable [imposition of intolerable management costs]

Part Two next week …