Some Thoughts on the Imperfection Theorems

For context, in A Quiet Revolution in Welfare Economics (Albert & Hahnel, 1990), the authors presented a welfare economics paradigm that was built off of endogenous preferences (determined within the economy) instead of exogenous preferences (determined outside the economy). Based off of this, they then give three imperfection theorems:

1. Markets are unable to properly price for public goods and goods with externalities in terms of social cost/benefit, which influences others to not buy them as they become too expensive. This leads them to become unviable as a solution to the free-rider problem.
2. Private enterprise will pay less for things that empower employees more and visa versa, which effects the bargaining power of labor.
3. Central planning is biased against self-managed work, supplying it less than socially optimal.

Now, these on their own are interesting criticisms to levy against each institution respectively. In my opinion, the thing that makes something more damning is something to back it up, so the first thing I ask when it comes to something like this is "can it be analyzed?" This isn't to say that it cannot be useful on its own, as leftism talks about theory quite a bit, but its always nice to demonstrate it in real life.

However, a recent (8/15/2021) email I sent that included this topic got an answer from Hahnel which left me intrigued (proper letter casing by me [originally in upper case], bolding by Hahnel):

I am unaware of any empirical work anyone has done about the imperfection theorems proved in Quiet Revolution In Welfare Economics (Princeton[...], 1990). That book was based largely on my doctoral dissertation at American University where I was a graduate student before becoming a faculty member. It was a critique of neoclassical welfare theory, [arguing] that we could learn a great deal of important things about different economic institutions and economic systems by modeling preferences as endogenous, rather than exogenous as they are traditionally treated in neoclassical theory.

So it appears that the book was meant mainly for theory and not for actual analysis. Not too great on my end.

However, that doesn't stop me and my hunger for a source or two; after all, the standard commodity was only a theory until people got to work.

---------------------------------------------------

So there are two general questions we can ask in order when trying to figure out the empirical reality of these theorems:

1. Is there any existing work that deal with each of the ideas that the theorems suggest?
2. What foundations can we set for any future analysis of these theorems that are in line with the original literature?

By this point, we turn into a literature overview.

------------------------------------------------------------------------

For theorem #1, we can find evidence in Nold (1992) that the demand for public goods can be higher than whats actually spent, including things like schools schools:

While the average current level of expenditure is equal to $1,814 per pupil, the average estimated demand is equal to $2,225 per pupil. In terms of public choice, this presents an interesting question of majority rule (median voter) versus efficiency.

Fernandes & Valente (2021) demonstrate that markets, faced with false claims, are unable to handle negative externalities from a consumer perspective:

...when false claims can be made, markets will appear very prosocial to the outside observer who will see widespread concerns for externalities and a price premium on allegedly ethical goods relative to conventional ones. In fact, conventional goods are just being falsely advertised as ethical. In addition, the price premium is seldom enough to cover the additional cost of producing a good that minimizes externalities. Even when credence claims are not allowed, the market will only partially internalize negative externalities, leaving thus room for some form of regulatory intervention.

Nawrot and Para (2014) show that in a case like cigarettes that even though they cause terrible externalities, they may still get consumed in large numbers:

Using of tobacco-related products is known for its detrimental influence on both individual and social health. Despite that, more than billion people smokes cigarettes worldwide. One of the main reasons is lack of perfect information and inability to deeply consider future health problems.

To top it all off, Templet (1994) also shows that externalities can cause some heavy damage for the economy in the form of "implicit subsidies":

The subsidies are generally supported by those less well off and the socioeconomic, energy and environmental indicators of the public sector worsen as externalities and subsidies increase.... Poverty, unemployment and income disparity are found to increase significantly while income, economic development attractiveness, energy efficiency and environmental quality decline with increasing externalization across states.

While it isn't much, it does give the road to proving the imperfection theorem for markets. Thus, it answers question #1 with a resounding yes.

For question #2, more research should either be found or done into analyzing the prices of goods relative to social costs and private costs. Along with this, it should be noted how much of an effect these prices, combined with preferences, have on future buying patterns. Also maybe some more writing on undersupplying public goods.

---------------------------------------------------------------------------------------

Theorem #2 can be extended to where we identify employee-empowering conditions:

[W]orking under conditions of (1) participatory decision making, (2) cooperation among employees, and (3) fairly perceived distribution of duties and rewards among employees, are all type-1 laboring activities, which are underpaid and undersupplied; whereas working under conditions of (4) wage discrimination, (5) employment discrimination, and (6) artificially created hierarchies are all type-2 laboring activities, which are overpaid and oversupplied.

For empirical evidence surrounding this topic, there's actually quite a lot.

In relation to (6), we can find that there's a heavy correlation between hierarchy power & income; the earliest comes from Simon (1957):

While the present theory is consistent with a positive correlation between compensation and ability, only an improbable coincidence would bring about equality between salaries determined by the mechanism described here and salaries determined by the marginal productivity mechanism. Hence, it would appear that the distribution of executive salaries is not unambiguously determined by economic forces, but is subject to modification through social processes that determine the relevant norms.

And the latest comes from Fix (2020), with empirical evidence to back it up:

I find that the relative income of CEOs increases with hierarchical power, as does the capitalist composition of this income. A model that extrapolates these data suggests that the trend among CEOs plausibly extends to the general US population. In other words, it is plausible that there is a three-way relation between income size, income composition, and hierarchical power in the United States.

And although these aren't the type-1 conditions that Albert & Hahnel originally described, it should be noted to see how something like profit sharing (which is in line with the trend of empowerment originally observed) can affect the employees.

Fang (2012) notes that firms that practice profit sharing increases the pay of employees:

On average, employees in Canadian establishments that adopted profit sharing during 1999-2001 appeared to benefit from the introduction of profit sharing, in terms of both their cash real earnings growth and total real earnings growth, in the five-year span following introduction of profit sharing. This advantage was both statistically and practically significant, adding about 15 percentage points to real employee earnings growth over the five-year period, a period during which employee earnings growth was generally modest.

Aerts et al (2013) shows that profit sharing at least increases product innovation in the firm:

Based on the matching results, profit sharing companies outperform non-profit sharing companies on both process and product innovativeness. However, according to the results of the conditional difference-in-differences method only product innovation is enhanced by the introduction of profit sharing. Hence firms introducing profit sharing are already more innovative with respect to both product and process innovations before they launch the variable incentive method of profit sharing. But profit sharing additionally has a positive effect on product innovations.

Florkowski & Schuster (1992) say that employees' answers to a survey show that there is support for it, and that it would increase organizational commitment:

One of the most significant findings of the present study is that companies can strengthen workforce commitment through profit sharing. The likelihood of doing so increases if plans are designed to meet certain expectations held by employees.

Coyle-Shapiro et al (2002), to back up the previous study with empirical evidence, also demonstrates that profit sharing increases organizational commitment:

Favorable perceptions of profit sharing served to increase organizational commitment while only organizational reciprocity predicted trust in management. The relationship between organizational reciprocity and commitment was partially mediated by trust in management.

And although it is still a bit, only 19-23% of firms utilize profit-sharing since 1963, which demonstrates that despite being beneficial for both the owner & the employee, its not as commonly used as it should be.

For future analysis, it might just be easier to say that more should be done into analyzing how employee-empowering practices get shoved aside in the traditional firm and the reasons for it. Along with this, it should be compared to how firm structures like worker coops compare to the traditional firm in this regard.

-----------------------

For theorem #3, its a bit difficult to quantify it for empirical purposes; its also difficult to find papers looking into the state of self-management under central planning. To try and deal with these issues, I'll focus on the USSR, as its the most popular example of central planning for the time being.

The Soviet system's methods of allocating labor were against self-management, plain & simple; according to Ian Jefferies' Socialist Economies and the Transition to the Market, the basic procedure for allocating labor went as such:

Market forces were heeded when the planners determined basic wage differentials, while the state controlled the education system, including the number of places available for particular courses of study. The state was able to use this combination, for example, to induce the increased supply of labour of the required skill needed to meet an altered production plan.

We can see immediately that the state used the education system to their advantage when preparing for their plan; an egregious violation of self-management.

And even though we can point to forced labor camps as a way that the state truly enforced what labor one does, their methods of centrally planning labor aside from it were also a bit of a mess.

High labour turnover during the 1930s, seen as a threat to plan fulfillment, was combated by means such as the ‘work book’. This was introduced in 1938 and held by the enterprise manager; without it a worker could not, in principle, find another job. Increasingly harsh legislation eventually made even absenteeism and lateness criminal offences. Graduates of universities and technical schools were assigned to a place of work for two or three years.

Trade unions were also controlled by the state, and prepared as such to where the worker & the manager were a part of the same union. Along with this, the usual functions of a union were contradicted by Soviet policy:

There was no collective bargaining between trade unions and management about basic wage and salary differentials, although the former exercised some marginal consultative roles. Strikes were considered  to be counter-revolutionary and in any case unnecessary, although they were not actually outlawed in the constitution. This reduced the role of trade unions to the transmission of party policies, ensuring favourable conditions for plan fulfillment, protecting workers’ interests (legal requirements as to health and safety, for example) and administering the social security system relating to sickness, work injury and pensions.

And without going in-depth, The Bolsheviks and Workers' Control by Maurice Briton covers the history of the Factory Committees, which advocated for workers' control and a rejection of state ownership, and how the USSR violently reacted to them. Need I mention Emma Goldman's recounting of how Lenin brushed aside free speech and free press and many more scenarios?

Overall, we see the silencing of anarchist press, the use of state education to influence others' decisions, and the absolute restriction of labor freedom. Self-management in the USSR had been thrown out in favor of the state in the god-forsaken name of socialism.

For further research, more should be written on the conditions of Soviet workers and how labor was allocated.

--------------------------------------------------------------------

After all this writing and reading through the literature, I can say that all of the imperfection theorems as given by Albert & Hahnel (1990) at the very least have evidence of empirical weight and more should be written on the validity of these. It seems likely that I will have to do this further research in the future, since A Quiet Revolution in Welfare Economics doesn't have much of a reputation outside the small circle of parecon, however if anyone else would like to write about this I invite them too.