Unpredictable Patterns #63: The price and cost of everything
Counterfactuals, fairness and the thickening narratives encoded in prices - and how you can use them to understand the future better
Dear reader,
Today the days are getting brighter and light stays longer. Summer is coming, and even if we live in a historically dark period there is always hope. Today’s note is on the theme of costs and prices - and how to think about them as general purpose tools for future studies and political analysis. I will confess up front to not being an economist - and there is a lot of theory about cost and price in economics - but I think these concepts are useful beyond that. I hope you find this interesting and agree!
On the last note - a thank you to reader A who suggested DAOs can be understood as a longing for a rational. apolitical organization - or at least an organization free from informal politics. This idea is interesting in that it suggests that if we could only fully formalize human interaction the resulting organization would be stronger, and so, conversely, informal politics makes us weaker. I am not sure that is true. I think the political problem is about living together, not solving problems - and that requires the messiness of the informal.
But I am often wrong! An interesting clue to pursue in any case!
Cost and price
Knowing the cost and price of things is an important piece of knowing the world we operate in. The cost of making something and the price it sells at is different, and therein lies a lot of the driving forces behind social change, human behavior and the future of our civilization.
The trivial example is the cost of producing a car and the price at which it sells. As we look closely at this, however, we see that it is not trivial at all - because the way we calculate cost may or may not cover the overall total cost of making that car for us as a society, not just for the company that produces it. There is a cost of making and driving cars that is not included in the original assessment, as well as a cost of living in a society that operates complex supply chains that give rise to geopolitical spheres of power - but these are not necessarily as well understood as the concrete costs of the materials and labor that go into the production of the car.
More and more often, politics will try to ensure that those costs are represented in the price in some way - as taxes, perhaps - and that the second and third order costs are not simply forgotten, but this is really hard and often an inexact science, since there are costs associated with not having effective transportation as well.
Economists have debated these issues long and well, and their understanding of cost and price is both deep and evolving. The political study of cost and price, however, is not as well developed. In a sense politics could well be described as a way to determine both the landscapes of costs and price in ways that create a community we believe we want. Sometimes it is helpful to approach a set of concepts so deeply entrenched as these in a naive way, and explore how we can use them to build our mental models anew.
Defense policy can be understood as the answer to a simple question about what the price should be for attacking us. A country without any possible defense is cheap to attack, whereas a country that has significant military capability, say, is much more expensive to attack. But defense policy should also be about the cost of defense, and the question of how we can generate the maximum possible price from the minimum possible cost. A quick analysis would suggest that the logical thing is to get nuclear weapons, since those weapons - credibly set up - create a near infinite price for the attacker. The challenge is, of course, that a deeper analysis shows that the cost of producing these weapons vastly misrepresents the civilizational cost of having them around. The very existence of these weapons increase the risk that civilization itself will end by a small, but significant enough percentage to make them very, very costly.
This raises an important question about cost, and that is the subject of cost. Who pays? That is a fundamental issue when we analyze costs and prices. A cost that is born by everyone generating a price that benefits but a few makes for a complicated political problem. Individual, group, national and civilizational costs are fundamentally different from a political perspective, and figuring out how to manage them is a core task for politics.
And how do we define these different costs? What is the cost structure of civilization? One possible answer is that it is the cost of managing the complexity generated by our consumption of energy (as pointed out by Joseph Tainter). Another is that it is the cost of constantly reducing every other individual’s cost of living and improving their quality of life - maybe it is the cost, then, of human progress.
What, then, is the price of civilization? Is it the individual contribution we are willing to make for living in one? This idea, that civilization has shared costs, but an individual price is intriguing. There are some goods that are only possible to buy individually, but can only be consumed collectively.
The distribution of costs is the distribution of power. If I eliminate all cost for you to do whatever you like, I have given you a lot of power.
The narrative singularity of price
Economics tells us that prices aggregate information and coordinate knowledge. A price is a uniquely dense point of knowledge that helps us understand how we value the world. Prices reflect data and knowledge, for sure, but they also reflect narratives. Robert Schiller and others have explored this idea of narrative economics, and they suggest that markets move with narratives more than anything else.
It is a point we have come back to, many times, in these notes: we understand the world only in narrative - not as data, information or even knowledge. We need to understand the world as knowledge mapped in time - this is the main thesis of a lot of French philosopher Paul Riceour’s work - and so if we unpack a price we should expect to be able to explore the narrative that leads to that particular price.
Markets are stories upon stories.
That is one of the reasons ”free” is such a problematic price. When tech companies argue that they offer their services for free they are making the point that the costs are absorbed elsewhere than in a direct price paid by the user, and that is trivially true. But since we have evolved to look for the narrative in a price, the idea that there is no price attached to what looks like a massive cost in producing these services produces a suspicious response.
The slogan ”if it is free, you are the product” is a narrative about a secret price, a price that is paid for you by others. It answers the basic question ”but where is the price attached to this cost?” that we have learned to ask by pointing at another transaction than the one where you would expect to find the price in the first place - the transaction between the advertiser and the company producing the services. It is one of the narrative complexities with many-sided markets that regularly gets under-appreciated.
Our ability to read prices - to unpack the narratives in them - is one key to trust. When we understand the costs that go into the production of something and the price, we also are able to assess the degree to which the company producing the goods or services profit from their interactions with us. It is the way we assess fairness. When we lose the ability to read a price, we also lose the ability to trust the company offering the service.
Pricing is how we signal fairness.
One example of this is how we tend to react to individual pricing algorithms. On the surface, such algorithms seem to make sense: a market should be looking at each individual transaction and seek to find a price point that represents the utility for the individual, after all. But we do not react well to knowing that we are seeing a price that was chosen uniquely on a set of assumptions about us. Early such experiments even forced the online retailers engaging in them to apologize.
We do not want prices to be narratives about us - there is a reason we call it ”price discrimination”. Increasingly, however, prices will be more and more adaptable. And if we are honest price discrimination has always existed - it is just that the resolution of price discrimination has been coarse grained and low.
That is likely to change as narratives compressed into prices are getting thicker. This thickening of prices means that they will take into account more and more variables and shift more than before, along lines that may be harder to predict. And prices can now reflect individual analysis at a much higher resolution than before.
Is that worse or better than than low resolution price discrimination available in different packages, bundles, programs and windows used today to shift prices?
Hal Varian, in the Handbook of Industrial Organization, pointed out something interesting in relationship to the coarse-grained often complex price discrimination we see today: he noted that it imposes a computational cost on the consumer who has to choose the right offering.
”One particularly interesting set of questions in this area that has received little attention concerns the computational costs involved in using complex forms of price discrimination. In the post-deregulation airline industry of the United States, airlines have taken to using very involved pricing schemes. Finding the most inexpensive feasible fight may involve a considerable expenditure of time and effort. What are the welfare consequences of this sort of price discrimination? Do firms appropriately take into account the computational externality imposed on their customers?
Even in more prosaic case of public utilities, pricing schedules have become so complex that households often make the “wrong” choice of telephone service or electricity use. Questions of simplicity and ease-of-use have not hitherto played a role in the positive and normative analysis of price discrimination. Perhaps this will serve as a fruitful area of investigation in future studies of price discrimination.”
Individually targeted prices could eliminate that, but would also require some new mechanism of legitimacy; a way to ensure that we still find the price fair.
Fairness in pricing is a strange thing. It mixes rational economic models with evolutionary psychology models that suggest we react with pain when we feel that someone cheats us. The most well-known example of this is the ultimatum game where one player gets to divide up 10 dollars and the other has to approve the bid or both lose money. A surprising amount of people go for 50/50, even though the rational choice would be for the proposer of the share to take almost everything - since whatever the decider gets is more than zero. Fairness, however, dictates that we reject offers that are wildly out of whack with what we expect.
If costs sets a lower bound for prices, fairness sets an upper bound. And that may be why we are so sensitive to individual price discrimination: we believe it is more likely for something to be fair if it applies equally to all - in fact, that is one component of evaluating fairness!
Unpacking the narrative in prices is a way to predict their future trajectories, but also to understand the driving forces behind them, and the scenarios that result.
Game theory and the future
Bruce Bueno de Mesquita, a political scientist engaged in understanding how game theory can be used to build robust predictive models for political systems, likes to point out the usefulness of studying history and politics with a very simple question. When looking at how people have acted in the past, you should always assume that they made the choices they thought were the least costly for them and generated the greatest benefit.
It seems trivial, even a little bit simplistic, but it is in fact a great tool in reconstructing other people’s world models.
The reconstruction of other people’s world models is key to predicting how they will act in the future, and it is also key to understanding how to change their minds. The idea of focusing on implied cost structures in individual decisions also ensures that we do not fall into the trap of just declaring the people we are engaging with as ”stupid” or ”clueless”. That assumption is just as bad as the assumption that they will have evaluated the situation according to our own yardstick for rationality.
What you really want to do is to reconstruct their rationality, the implied cost structure and the benefits they must have seen in order for them to make the decision they made.
This is a useful tool in competitive intelligence, as well as in political analysis. Look at what other companies in your sector are doing and assume they are making choices that create the least costs and the most benefits - what must then be true about their worldview if you back it out from their actions?
We can not assume everyone is rational in the same way. But we can assume that people are consistent, and hence that their actions will tell a story about their assessments of the choices they had to make.
Importantly, Bueno de Mesquita suggests, the actions not taken were thought to have a higher price than the actions taken. What does that mean? If you unpack the narratives in those prices - what made them higher than the price for the action ultimately chosen? This method - the careful and analytical use of the historical counterfactual - is sorely underestimated.
Buena de Mesquita’s models are interesting in that they do this for n-person games, and then use the resulting models to predict political choices in ways that have proven to have an interesting track record. Using counterfactuals unpacked from assumed prices extracted from the actions taken is a good way to explore future possibilities:
Game theory provides a useful way to structure counterfactual arguments in that the solution to extensive form games requires explicit attentiveness to counterfactuals in at least two central ways: the solutions or predictions from extensive form games depend on what is expected to happen “off the equilibrium path”; and games often have more than one equilibrium solution, each of which represents a plausible state of the relevant world. Equilibria not chosen in a game with multiple equilibria and off-the-equilibrium-path expectations represent two important sources of counterfactual argument that influence the chain of causality from a game-theoretic perspective. For a game theorist, therefore, counterfactuals are not to be avoided, but rather to be used as important tools in understanding reality and beliefs about it.
(From ”Counterfactuals and International Affairs: Some Insights from Game Theory”)
Every price casts a shadow of counterfactuals.
The price of political choices is votes
To at least some degree, power is about the ability to impose costs. If the powerless can impose no costs at all, the and the powerful may even impose the ultimate costs of death, then political power seems to be about figuring out who can force which costs on who.
This is a simplification, but a helpful one. It points us in the direction of an interesting exploration of the sources of political cost. Which costs matter in politics?
One way to answer that question is to say that all political choices ultimately are priced in votes. The price of a decision is the votes it will cost you. Any politician worried about re-election will pause if faced with a choice that would lose them 50% of their votes.
In the simplest possible model of political choice, then, we assume that politicians will vote for a proposal if they think it will at least not cost them any votes (all things equal that means they stay in power - this is about curating your wider, already acquired base).
Ideally, they would like to gain some votes, but gaining votes is harder than losing them - and is probably often is the consequence of a pattern of choices than a single decisive choice. Gaining votes is about convincing swing voters - and the nature of the swing voters is that they swing if two things are true: their old candidate has made choices that cost them their confidence, and the new candidate can exhibit a pattern or characteristics that makes them more attractive than last time.
In this model, then, the great political game is about creating costs for politicians, ensuring that the decisions you do not want cost them voters, and that the decisions you want them to make costs nothing and contribute to a pattern that can win the swing vote.
This is a brutally transactional model, but there is some truth to it.
Modeling the choices that decision makers have made so far, then, is a good way to get a sense of how they view their cost structure. What bet are they placing in terms of the choices they make? How does that suggest they view the price in terms of votes lost in different issues?
Next comes the question of how to change that calculus.
There are two major dimensions of political cost that matter here. One is the imposition of cost in an established issue - suggesting that the choice before the decision maker is a familiar choice where they know what the equation looks like, and where they should vote the way you want because that is what minimizes the price in votes. The argument you have to make in this model is simply: this is like that, and you already know you need to vote for that.
The second dimension is the creation of a new kind of cost. This is tricker, but requires showing that there are voter groups and segments that care enough about an issue to make it an issue that the decision maker can lose voters on. This is the creation of a price for a vote that might otherwise have been seen as free.
Examples of this abound: political organization is largely about the creation of prices for votes that would otherwise have been free (in terms of the votes lost or possibly gained over time). An early example in tech was the organization of sellers on the platform and the way they then let their politicians know that they cared about different Internet policy issues. The creation of a new price tag on issues is long term work, but fundamentally important - especially in the tech sector, where new issues come up all the time. This is interesting: the way a new entrant or challenger usually approaches politics is through thinking that they just need to be nice and show that they have a general benefit for the voters - but the reality is that the incumbent play will be to show that there are many specific disbenefits and so the newcomers need to be ”reigned in”.
The struggle between incumbents and new entrants is over the power to set the price tags on the decisions that matter to them - and the incumbents have a clear advantage in this fight.
*
So what?
Cost and price are well known concepts, and the use of them in economic analysis is well understood. They are, however, broader concepts too, and can be used in political analysis and future studies in a number of interesting ways. A few things to try out:
Use implied costs and prices for different historical scenarios to construct counterfactuals and explore them. Which changes in costs or prices would have led to another outcome - how could those changes have been accomplished?
Figure out who sets the price tag on a decision, and if you can how to change that. If you set the price for decisions, you indirectly will be able to influence them. And if you realize you cannot, the best use of your time is to mitigate the results.
Explore the fairness boundary of the prices in your work. How are prices set, and what narratives do they encode? Those narratives are getting thicker, so understanding them will be harder and harder - but also more important. Unpack prices into the stories they tell.
Finally, then, are there things that should have no price? Naturally. But all things have a cost.
N