Gouging?

Almost all of the attention surrounding the rising costs of energy has been directed at Big Oil, and the price rises are attributed to their greed. Were they less greedy in the past, when prices were lower?

But what about Big Sun? According to the experts over at Solar Buzz, the price of Photovoltaic modules was decreasing until June 2004, and has been rising since (see image and their extensive statistics).














How does that compare to what oil has been doing?













It looks like prices were steady on the oil front until late 2003, and decreasing in the PV industry until June 2004, and since then both have been climbing. There are at least two ways to look at it:

• Bad: Energy costs impact people of limited means (poor people, retired people on fixed incomes) hardest.
• Good: Rising energy costs mean more conservation, more searching for oil, more searching for alternatives.
  

Which one you favor or emphasize depends on whether you more heavily weigh the short or long term, respectively. You might also quibble with how Solar Buzz creates their statistics, but I think the overall trend is sound. Whether you invest in oil production or solar production, profits are improving in both sectors and look to continue doing so for the foreseeable future. Given the situation in the world, and the fact that we are not nearly ready for the 100% (or even the 5%) solar solution, oil looks to be the stronger investment opportunity for some time to come.

Technorati tags: gouging energy

Communal-emh

Communal-emh -- where "emh" stands for "every man for himself" -- is the first of Oliver Williamson's modes of organization that comes under the category of "collective ownership". In this system, all of the stations in a factory are owned collectively, but laborers move their raw and intermediate materials from station to station themselves, then sell the finished products on the open market. This another cooperative system (the Federated being the other covered so far). OEW seems to include this as an intermediate step between the craft production (black-smithing, where one man built everything by himself in a simple process) and the next step in communal organization, Peer Groups.

• It is classified by OEW as an "collective ownership mode" (as opposed to entrepreneurial or capitalist)
• Compatible with agoric modes, in my opinion
  
• Although there isn't much contracting involved, OEW puts it under "periodic contracting" to handle situations that might arise from temporary disability of workers (where intermediate products might otherwise stand idle)
  
• In terms of contractual hierarchy, OEW classifies it as being low. There is no central agent, and workers own their intermediate products.
  
• In terms of decision-making hierarchy, again with no boss, it is also at the lowest end of the scale
  

So, life the Federated system, Communal-emh is not very hierarchical. Of the six organizational modes, Communal-emh is the least efficient.

Efficiencies:
a. Product flow
1. Transportation expense - Having everyone under the same roof means that transportation costs are economized.
2. Buffer inventories - Since you may go much faster than other workers, and therefore have to spend downtime waiting, it will be worthwhile to process in batches. This would be a waste compared to a system that moves material through in a constant flow.
3. Interface leakage - The fact that you are carrying your own raw and intermediate products means you have an incentive to economize on leakage.

b. Assignment attributes
4. Station assignments - All people are not going to be equally good at all jobs, so this method does poorly at economizing on station assignments.
5. Leadership - With regard to leadership, Williamson rates Communal-emh as economizing, but no explanation is offered.
6. Contracting - Communal-emh is rated poorly for contracting for reasons given previously.

c. Incentive attributes
7. Work intensity - This mode will economize on work intensity because it is, after all, emh -- every man for himself.
8. Equipment utilization - This mode will economize poorly on equipment utilization. The equipment is owned in common and used by everyone, so care of the equipment becomes a classic Tragedy of the Commons problem.
9. Local shock responsiveness - Should a worker fall ill or fail to come to work, that doesn't stop everyone else, but it does substantively halt that worker's output.
10. Local innovation - OEW rates this mode low for local innovation with no explanation. If you want to change a process (improve a machine, build new dies, invest in new equipment), you have to obtain consensus of everyone. As most people will realize, people have different responses to change, so this would be cumbersome.
11. System responsiveness - This is effectively a macro view of local innovation. If the market changes, and requires a systemic change to adapt to new conditions, the requirement of consensus is multiplied many times. Thus, the Communal-emh system is far more difficult to change on a large scale than other modes.

Bottom line: minimally hierarchical, but the lowest efficiency score.

BTW, yes, some of these seem like rent posts, but my choice was either to detail and summarize all at once, or to detail one at a time and then summarize. I'm getting there, but slowly.

Technorati tags: organization theory "Oliver E. Williamson" "Transaction Cost Economics" book review economics

Federated

The second of Oliver Williamson's (OEW) organization methods from The Economic Institutions of Capitalism (described in this post) is another "entrepreneurial" mode he calls "Federated". He describes it thus:

Stations are located side by side in a common facility. Intermediate product is transferred across stages according to contract. So as to avoid the need for supervision orcontinuous coordination, buffer inventories are introduced at each station. Subject to the condition that buffer inventories do not fall below prescribed levels, in which event penalties are assessed, each worker proceeds at his own pace.

He goes on to say that it is "uncertain and perhaps doubtful" that this mode was ever widely used. He cites Pat Hudson ("Proto-industrialization: The case of the West Riding World Textile Industry in the 18th and early 19th centuries," History Workshop, 12 (Autumn, 1981)) as saying that "the majority of early woolen mills were occupied and run [...] by small manufacturers ... rather than by wealthy mercantile concerns," but that it is not clear whether they traded between stations. But if the mode didn't exist, why bring it up at all? OEW explains that it is useful as a hypothetical, and especially as one in which worker autonomy is preserved.

So what are the salient features of the federated system?

• It is classified by OEW as an "entrepreneurial mode" (again, as opposed to collective or capitalist)
• Compatible with agoric modes, again in my opinion
  
• OEW considered it to be a "continuous contracting" mode
  
• In terms of contractual hierarchy, OEW classifies it as being low. Workers own their own equipment, and there is no central agent
• In terms of decision-making hierarchy, with no boss, it is also at the lowest end of the scale
  

So it's not very hierarchical, but how does it stack up efficiency-wise? Poorly, as it turns out.

Efficiencies:
a. Product flow
1. Transportation expense - Having everyone under the same roof means that transportation costs are economized, an improvement on Putting out.
2. Buffer inventories - The fact that everyone is dependent on upstream contractors puts you at their mercy. If you happen to be downstream of a poor operator, so much the worse for you, and even contractual countermeasures aren't going to help very much. Buffer inventories are anathema to flow production.
3. Interface leakage - This also receives low marks (recall that this means loss of material from one station to the next).

b. Assignment attributes
4. Station assignments - Same as Putting out (good).
5. Leadership - In this category, the federated system loses marks. There is no central figure to point out when prices are too high, manufacturing processes are taking too long, or whether the group is in fact making the right product.
6. Contracting - Same as Putting out in OEW's opinion, but I suspect the same moderating tendencies as before. That is, the use of modern predictive maintenance practices make this less important, but the comparative cost of contracting between this and systems which economize on contracting will tip in their favor.

c. Incentive attributes
7. Work intensity - Same as Putting out.
8. Equipment utilization - Same as Putting out.
9. Local shock responsiveness - Same as Putting out, or perhaps worse. At least with the central agent doing the subcontracting in putting out, someone might recognize the problems and redirect work to another subcontractor.
10. Local innovation - Same as Putting Out.
11. System responsiveness - Same as Putting out.

Bottom line: minimally hierarchical, but the same 5 out of 11 on efficiency. This is one of the three modes of organization that could be roughly described as "cooperative".

Technorati tags: organization theory "Oliver E. Williamson" "Transaction Cost Economics" book review economics

Putting out

Continuing from Flow:

The first production mode listed by Williamson (OEW) in The Economic Institutions of Capitalism is "putting out".

• It is classified by OEW as an "entrepreneurial mode" (as opposed to collective or capitalist)
• Compatible with agoric modes in my opinion
• Usage in early gunmaking industry - David Hounshell writes, "[In 1854, t]he British obtained their military small arms through a system of contracting with private manufacturers located principally in the Birmingham and London areas.... Although significant variation occurred, almost all of the contractors manufactured parts or fitted them through a highly decentralized, putting-out process using small workshops and highly skilled labor. In small arms making as in lock production, the 'workshop system' rather than the 'factory system' was the rule." (Hounshell 1984, p.17)
  
• OEW considered it to be a "continuous contracting" mode
• In terms of contractual hierarchy, OEW classifies it in the middle of the pack. Workers own their own equipment, but there is a central agentIn terms of decision-making hierarchy, with no boss except the central agent, it is also in the middle of the pack
  

This mode is what is usually referenced in works like Malone's The Future of Work, though not by this name, as the work organization of the future. Crowdsourcing is the completion of projects such as wikipedia and linux that are "put out" to many different contributors. Linux particularly is relevant to this since Linus Torvalds was the central agent in the early days. I also think that putting out is essentially the same as outsourcing, so this has more relevance to modern manufacturing methods than I think Williamson's 1985 treatment would have indicated.

With respect to Williamson's efficiency analysis, however, it doesn't do well.

Efficiencies:
a. Product flow
1. Transportation expense - not good. Intermediate pieces have to be transported between either the put-outees and the central agent, or sequentially between each other. As Malone points out, the decreasing costs of communication brought by e-mail and the internet have rendered transportation costs trivial for purely intellectual pursuits such as writing. Indeed, I am writing this with Writely, which is expressly designed for collaborative writing. The Economist recently highlighted a directly relevant and concrete proof of this point, which Greg Mankiw commented on: Adobe's PDF has made bicycle messengering an endangered species. The expense of outsourcing to Asia suffers primarily from this inefficiency, but not so much as to overcome other efficiencies. One such possibility is economy of scale, since Asian (especially Chinese) factories can produce for domestic consumption as well as North American and European consumption.

2. Buffer inventories - the frequent travel and dependencies of one process on another requires buffering and batching, and so makes the traditional putting out system very ineffective in terms of economizing on buffer inventories. That is a major difference between traditional putting out and supplier-managed inventory subcontracting as used by Toyota and others.

3. Interface leakage - The opportunity to blame material usage inefficiency on the preceding or following processor and to embezzle or lose material at the interfaces between processes. Consider what this means for outsourcing: if products are merely coming to the factory, the outside source will manage the materials much better than if sub assemblies were going from the factory out, or between outside sources. I understand this is the case for some subassemblies in automobiles, where for example parts will go from one subcontractor to another for integration into a harness, and then to another for integration into the vehicle. How are such discrepancies handled? Such leakage was reportedly bad in the textile industries in the 18th and 19th centuries; I'm not sure how bad it is in textile manufacturing outsourced to Asia and Latin America.

b. Assignment attributes
4. Station assignments - Putting out economizes well with station assignments. If someone is unable to do the work, he is not picked up in the next contract or cannot make a profit and takes himself out. That was especially true in the system as practiced in the 18th and 19th centuries, where work was put out to family-based contractors. That's not so clear when outsourcing to modern contractors.

5. Leadership - Putting out economizes well on leadership for the same reasons.

6. Contracting - Putting out does not lend itself to other contracting arrangements in Williamson's analysis. I think it may be a mixed bag when comparing to modern subcontracting. Williamson uses the example of contracting with maintenance specialists. TPM suggests that the operators learn to do the maintenance, but does allow for specialists to help with the harder maintenance. One could also expect that contract costs would reflect whomever was better at maintaining the machines. However, while the requirement for maintenance might be a very minor cost, contracting may make a difference when you are looking at marginal advantages between one mode and another.

c. Incentive attributes
7. Work intensity - Putting out is highly correlated to work intensity. For a one-person operation, the outside source is essentially making piece rate. For a multiple person operation as was typical in the 18th century when people took home work in which the whole family participated, it was easy to see who was shirking and to shame them into keeping up.

8. Equipment utilization - Since the outside contractors own their own equipment, they are likely to economize on equipment utilization and not abuse it.

9. Local shock responsiveness - The outside source was not economical with respect to such local shocks as illness, injury, or machine breakdown. Such shocks comparatively favored those operations where everything was together under one roof and workers had incentives to back each other up (or at least did not have disincentives to doing so). Look at what happened to Ericsson ($) when a semiconductor fire destroyed the production capacity at a subcontractor in New Mexico: they ended up leaving the handset business.

10. Local innovation - As with equipment utilization, putting out benefited from those contractors who were able to both innovate and to keep their innovations to themselves because they were not working under a roof with everyone else. Local innovation is a requirement in order to be a subcontractor to Toyota.

11. System responsiveness - This efficiency is not well defined, but basically describes how well the system economized in response to outside shocks or changes in the marketplace. Although Williamson scores this one low, it is not clear why. Putting out was common when laborers were skilled craftsman and every product was unique. Thus, mass customization was the norm, not a new marketing strategy. It could be argued that putting out was obviously unresponsive to system shocks because it eventually gave way to new methods of organization, but given that outsourcing is commonplace and my assertion that outsourcing is the new putting out, that isn't true either. It could be that Williamson scores it as not being responsive because in comparison to other modes the "system" of putting out couldn't turn on a dime. Toyota's experience with the 1997 Aisin fire indicates the opposite, however. As described in Jeffrey Liker's The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer, Toyota has such a strong relationship with its suppliers that they recovered within 24 hours of the fire. Compare that with Ericsson's failure!

Although not covered in these areas of efficiency, Quality Control is not explicitly addressed. In might be implicit in local innovation, but they might be more inclined to innovate with respect to labor cost reduction than in making sure that components and subassemblies built by other putting out contractors fit together. In fact, that may be part of the problem with system responsiveness: given that one subassembly manufacturer might understand how to innovate and economize on the whole by changing his subassembly, but that would require an interface change and a diseconomy by another subcontractor, how do the three parties (central agent and the two subcontractors) divide the savings? There is a cost savings or quality improvement that can be passed on to the customer at either a gain in per-unit profits or market share, so everyone stands to benefit, but the complexity of the contractual arrangement needed to realize that gain may be so difficult that it isn't made and nobody gains. This is a market failure for which I know of no formal name, but for which the Second Best solution is to change organizational structure. It arises from the fact that it is difficult to say from which subcontractor the improvement comes: he who thought of it and stands to profit from his contract arrangement, or he who has to implement it and stands to lose from his contract arrangement.

Bottom line: Williamson finds putting out to be only moderately hierarchical, but not very efficient. It succeeds in only 5 of the 11 areas of efficiency. However, if subcontracting is the same as putting out, one must wonder what Japanese companies are doing since they are apparently using more subcontracting now than in the 1980s (damned if I can remember where I saw that statistic, and note that it was "Japanese companies", not Toyota). My guess is that the TWI methods now known as kaizen and the JIT methods have shifted how putting-out works sufficiently to overcome the inefficiencies noted by OEW.

Flow

In Lean Thinking, the authors reference a book by Mihalyi Csikzentmihalyi called Flow: The Psychology of Optimal Experience. I came across it again in a series of comments and posts on Catallarchy, so I finally broke down and bought it.

I heartily endorse the book. [Update: I confess to having been overexuberant about it. I think it was a case of extreme confirmation bias. In retrospect, it was neither a great read nor breathtakingly original. As I recall, the first third was really slow and promised more than the last 2/3 delivered.] The author (whose name I remember by pontificating that what this nation needs is a good "Six Cent Tamale") has conducted research at the University of Chicago for decades, primarily by giving people beepers that go off at random intervals, at which times they answer a series of questions about what they are currently doing and what their mood is (gross oversimplification, but that's what you get from reading psychology on a blog kept by an engineer). His conclusion based on this and other data is that people are happiest when they are involved in something that is so challenging as to require their undivided attention, but not so challenging as to be impossible to complete. In other words, people should have goals that are attainable, but only just attainable. In practice, that means that many people are just as likely to be enjoying themselves at work as they are outside of work; in many cases more so. Outside of work, many people are unable to find something that keeps them challenged. Our leisure activities are too dominated by television and other unchallenging diversions and his research showed that people were not as happy doing those things as they thought they might be (You can regularly see "happiness research" debated on econlog, The Fly Bottle, and agoraphilia).

In my last post on the new workplace, I surveyed three books that discuss the changing organizational structural (The 5th Discipline, The Future of Work, and Markets in the Firm), but didn't get around to pointing out what I would like work to be like. It seems to me that purging the workplace of Taylorism is essential to balancing the seemingly conflicting goals of making work for workers enjoyable yet productive for their employers and/or customers. Taylorism can best be expressed in the statement, "Managers think, workers do," a clear formula for thwarting the "flow experience" described by Csikzentmihalyi as essential to happiness. Both W. Edwards Deming and the more recent Lean or Toyota production literature emphasize the need to banish the 8th waste, that of neglecting the creativity of workers, in order to improve quality, reduce costs, and increase profits. In Markets in the Firm, Cowen and Parker bring forth the quote that scared much of America in the 1980's, Kanosuke Matsushita's famous claim that "Yes, we will win and you will lose. For you are not able to rid your minds of the obsolete Taylorism that we never had." It appears then that the answer to improving the job for both workers and management is to challenge workers with puzzles that they can solve. The means by which this is done is the self-directed continuous improvement team concept (call it what you will - Quality Circles, TQM, whatever - as long as you do it).

Fighting Taylorism is frequently cited as one purpose for unions. However, union positions on both Taylorism and self-directed teams have been mixed. UAW chief Walter Reuther bought into Taylorism and sought only to maximize their bargaining position on economic considerations rather than on the work experience. But they aren't embracing the new paradigm, either: Stephen Bainbridge (popularly known as Professor Bainbridge) has pointed out that workers tend to favor hierarchy over self-directed production contracts. In this article, he says,

As John Witte contended in his important case study of self-directed work teams, "there is little reason to suspect that most workers would either endorse participation or become actively involved if the opportunity arose." To the contrary, Witte posits that workers generally accept hierarchical authority and perceive obedience to authority as an integral part of their job: "for the majority, disobedience is unthinkable."

Witte's pessimistic prediction is confirmed by a wealth of data. Only about half of the participants in Witte's own case study were willing to change jobs in order to get more participation and that rate declined rapidly if doing so would require longer hours or lower pay. A study of transportation firms found that long-term use of employee involvement initiatives increased stress and decreased employee fulfillment. [Anderson (1996).] A study of "empowered" employees versus a control group of unempowered employees within a single insurance company found no statistical difference between the two groups on such productivity-related issues as motivation or even on some job satisfaction measurements. [Thorlakson & Murray (1996).]

And at the website of the United Electrical, Radio and Machine Workers of America (UE), where they claim that "Most of these programs [kaizen, QC, and Just in Time] have evolved from the work of Frederick Winslow Taylor, who is credited with developing 'scientific management,'" they dedicate two pages (starting here) to throwing FUD at such management "schemes", among which there are two types:

One type strives to get the workers, with or without the union, to act in groups to help management improve productivity. These are sort of brainwashing sessions. They strive to get workers to think like a boss and to come up with ideas on how to cut other workers, speed-up production and ways to do more work. Quality Circles, Team Concept, and Kaizan [sic] fit into this category. These schemes often try to undermine the union by setting up non-union committees that will "make decisions for the workers."

A second employer scheme is usually tied to changes in how production or the work being done is organized. This could be cell manufacturing, Just in Time, Direct Flow Technology, pay for knowledge etc. These changes can have a direct effect on parts of the contract. Seniority, layoff and recall procedures, pay grades, and transfers may all be affected. This also gives the union some power, because in most cases the employer cannot make changes to the contract without the union membership's approval. [all emphasis in the original]

In other words, it appears that the ongoing demise of the Sloan/Taylorist management method in favor of the Toyota system, which achieves flow in both the production process and in each workers' experience, is coinciding with the demise of the unions in spite of them, not because of them. (Still, I caution readers to consider the reasons cited in the Bainbridge article: some worker resistance to self-direction may be understandable).

Since completing Flow, I started reading From the American System to Mass Production, 1800-1932 : The Development of Manufacturing Technology in the United States by David Hounshell, which I first saw referenced in the other Womack, et al book, The Machine that Changed the World. I was immediately struck by how Ford employees referred to their processes as "flow".

William Klann, a Ford deputy who was deeply involved in the innovation [the assembly line], agreed but noted that an equally important source of inspiration was flour milling technology as practiced in Minnesota. Klann summarized this technology in the expression "flow production."

So far, I find this book to be full of surprises, having already exploded the myth that Eli Whitney developed the first musket with interchangeable parts, and having revealed that the idea came from a French general through both Jefferson and a lieutenant of Lafayette who also happened to propose the idea for West Point.

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

Are hierarchy and team production the only two possibilities for organizing labor? What about the old methods of putting-out and inside contracting? What about communally owned facilities?

Oliver Williamson attempts to address those questions in The Economic Institutions of Capitalism. Though the analysis is ultimately unsatisfying, for reasons explained by Williamson himself (and to which I will return in a later post), it is an interesting start and at least as interesting as the other books surveyed in my last post (but substantially more difficult reading). Williamson analyzes 6 different schemes on 11 different criteria (listed below in his notation style) and determines that hierarchy is the best system in terms of efficiency. In my next few posts, I propose to relate this analysis with comment.

Production modes:

• Putting-out
• Federated
• Communal-emh (every man for himself)
• Peer groups
• Inside contracting
• Authority relation (hierarchy)
  

Efficiencies:

a. Product flow

1. Transportation expense
2. Buffer inventories
3. Interface leakage

b. Assignment attributes

4. Station assignments
5. Leadership
6. Contracting

c. Incentive attributes

7. Work intensity
8. Equipment utilization
9. Local shock responsiveness
10. Local innovation
11. System responsiveness

The New Workplace

I just finished reading Thomas Malone's The Future of Work: How the New Order of Business Will Shape Your Organization, Your Management Style and Your Life (2004) on the basis of this review at the Austrian Economists blog, and because I've been searching for answers to questions that Malone purports to answer. The same review references a paper written by Tyler Cowen and David Parker called Markets in the Firm: a Market-Process Approach to Management (1997), which I read today. Of the two, Markets in the Firm was a much better value, but it still didn't answer my questions.

Malone is a professor at MIT, the same place Womack, Jones, and the Lean Manufacturing cheerleaders call home. It is also the home of Peter Senge, whose 5th Discipline (1990) was a disappointment. The 5 disciplines referenced are:

1. Building Shared Vision
2. Mental models
3. Team Learning
4. Personal Mastery
5. Systems Thinking

If he had simply done a better job with the systems thinking and then referred readers to Stephen Covey's The 7 Habits of Highly Effective People (also written in the same year), I would have been more satisfied. Systems Thinking is old school to anyone with an undergraduate degree in engineering (especially electrical), but Senge tries to give the impression that this is something really new and radical. It could be, but his explanations and simplistic diagrams leave much to be desired. I think a great deal of the emphasis on systems thinking comes from MIT's past success with this theory: MIT researchers made a lot of press with their application of Systems Thinking in Limits to Growth and Senge quotes LtG co-author Donella Meadows extensively.

The rest of the book left me flat. I felt that Covey does a better job of describing how to build a vision and how to achieve personal mastery. Developing mental models and performing team learning mostly involve getting your assumptions out on the table and then building extensive computer programs around your processes. Both of those require skilled facilitators and model builders - since most of us don't have access to Senge and his associates for free, I think it's safe to assume that we can't access the methods described in the book.

Malone, on the other hand, stretches what should have been a 20 page magazine article in HBR into a 175 page book. Yes, the cost of communication is getting cheaper. Yes, you will be able to communicate laterally. Yes, that means that knowledge workers will be able to work in quasi-craft production groups. According to David Friedman (via Roderick Long and the Multualist Blog, there is a word for it: agoric. But after telling us that it is happening, and giving a few examples, the advice was rather sparse. As a manager, if you have the right type of firm (not clearly specified), you will have to give up command and control and instead use cultivation and coordination. What will you cultivate? Good practices. What will you coordinate? Profitable activity.

Malone specifically talks about using outsourcing so that the firm will simply be a center about which creative, productive, and profitable activity will take place. He uses Wikipedia and linux as examples, but ... Wikipedia is not done for profit, and linux has only been profitable for people who were in early. At some point, if you are in the business of selling things, someone actually has to make them. And if they want to get paid for it, they can't give them away for free.

The Cowen and Parker article was the most satisfying of the three. For one thing, it summarizes recent economic theorizing about the nature of the firm in a short section. This includes Transaction Cost Analysis, or New Institutional Economics, of which Oliver Williamson's work is the best known. For people interested in it but who don't want to tackle The Economic Institutions of Capitalism, this is the way to go. They trace the changes in industrial organization from craft production through the Industrial Revolution, Taylorism, and into the modern age. They declare Taylorism to be dead, killed by the intense competition brought by globalism.

The goal of their paper is to introduce the idea that firms are similar to states, Taylorism is analogous to central planning, and modern organizations are starting to look more like a market economy. This was obviously appealing to me in light of this post I wrote several months ago. I wrote that Lean production was Hayekian in that it allows people to share information and organizations to learn more quickly. In additioin to emphasizing that principle, Cowen and Parker write that firms should try to be more market like by design by using market-like incentives and clarifying property rights and responsibilities within the firm. They use Koch Industries as an example; Koch actually trademarked the phrase Market-Based Management (R). As the book he co-authored with Norm Bodek seemed especially concerned with the comparative accounting practices of GM (and especially Donaldson Brown) and Toyota, Bill Waddell should be especially interested in one of their closing comments:

This paper has considered some general principles which are a guide to how market economics can aid management. Future research needs to focus on the internal and institutional impediments to the use of these principles. One particular area that needs exploring is current accounting practices. The development of 'activity-based accounting' appears to be a step in the right direction by allocating joint costs or overheads more effectively so as to identify the true costs of production in various parts of the firm. [emphasis added]

Still, I am dissatisfied with the general sharing of information between management science books like 5th Discipline and The Future of Work and economic theory. Cowen and Parker mention W. Edwards Deming, but don't elaborate. Markets in the Firm was written in 1997, so they would have had access to many of the Lean canon (they cite The Machine that Changed the World in the bibliography), but don't explore much of it. If an article mapping Lean methods to transaction cost economies hasn't been written, it needs to be.

The purpose and benefit of unions

I have moved from being a non-exempt employee in a union shop to an exempt employee who works with unionized employees. Over that time, my thinking about unions has shifted to -- believe it or not -- one that is more friendly toward unions. More recently, I made my way through Oliver Williamson's The Economic Institutions of Capitalism, where I got an even wider perspective. Williamson discusses three facets of unions (Monopoly, Efficiency, Voice), among which efficiency (agency and governance) is the one which coalesced some recent thoughts I have had on the subject. I cover here four of those areas:

• Monopoly: The union restricts the availability of labor by keeping the company from going out to find competitive providers of labor. The resulting artificial scarcity is used to increase the price of labor. Also called rent-seeking.
  
• Power: The union provides additional bargaining power to the individual, balancing the unfair advantage held by the firm. This is the unions' rhetorical perspective, and one that Mark Graban seems to have bought into in this post considering his choice of one-sided support for the position. Bill Waddell's explanation of the situation in China is much more interesting and congruent with my own views on sweatshops.
• Governance: The union agreement is used to protect both the company and the individual, both of whom invest in specific assets. The company trains workers in its unique processes and neither desires early quits nor excessively harsh or arbitrary discipline by maverick supervisors, while the workers want protection from expropriation. CBA is a constitutional arrangement for a system where management and workers figure out the details for themselves. This is clearly the focus of Williamson's analysis.
• Agency: The union is a convenient way of aggregating opinion, developing consensus, and communicating with management more effectively. This is a transaction cost reduction.

One question raised in EIC was why Collective Bargaining Agreements (CBAs) are subject to frequent renegotiation.

1. One reason is the National Labor Relations Act (NLRA or Wagner), which allows decertification in only two circumstances: the first is if a contract is not in place, the second if a contract has gone on for more than three years. During decertification, a competing union could organize the workers, so it is in the interest of the existing union (though not necessarily the workers) to make contracts that last no more than three years.
2. Another possibility is that the parties are not privy to complete information about the future, consistent with Williamson's governance model. I agree with this, also, since most of the adjustments to our contracts serve to clarify language that resulted in a disagreement in the previous contract.
3. The truth probably includes those reasons and more. We might expect to see longer term contracts in stable industries (railroads), and shorter contracts in volatile industries (airlines) with built-in COLA adjustments.

In spite of these rather positive views, though, the unions themselves continue to push "us against them", class war mentality. The idea of union as a necessary evil to stand against corporate abuse is probably not valid today. To believe it, you have to believe that the government is the worst employer there is today since that is where unionization is concentrated and increasing. You also have to believe that management thinks they will get more out of miserable employees everywhere and always, and that no good innovations take place in the absence of government and unions. You only have to know about people like Robert Owen, Henry Ford, Eiji Toyoda, and J. J. Hill to know that's not true. In fact, in this story [hattip: CafeHayek] in the NYT about minimum wage activism, author Jon Gertner point out that the business owners "were not unsympathetic":

[T]he restaurateurs who took the stand, like Rob Day or Elizabeth Draiscol, who runs the popular Zia Diner in town, opened their books to show that their margins were thin, their costs high, their payrolls large. They cared about their employees (providing health care and benefits), trained unskilled workers who spoke little or no English, gave regular raises and paid starting salaries well above $5.15. They had built up their businesses through an extraordinary amount of hard work. Draiscol testified that her restaurant, for instance, had $2.17 million in annual revenue in the fiscal year of 2003. Though her assets were substantial - a restaurant can be valued at anywhere from 30 to 70 percent of its annual revenues, and Draiscol said that Zia had been appraised at 66 percent of revenues, or about $1.4 million - she earned a salary of $49,000 a year. [emphasis added - $49,000 a year is about $23/hour for a 40 hour/week job, or about $12/hour for the average small business operator. It does not account for capital appreciation of the business, though.]

Williamson also makes a strong case for his governance theory by looking at the order of unionization. In his efficiency explanation, unions serve to protect both the employees and employers in a bilateral relationship where each has a vested interest in maintaining firm-specific skills. He observes that the first successful unions were in those industries where workers developed specific skills, like railroads. Unionization occurred much more slowly in areas where workers have no specific skills (agriculture) where they can (relatively) easily find new employers and employers can easily find replacements. He points out that industrial unions have been successful only in post-Wagner environment, but that class-based unions would require much greater legislative interference to survive.

One last point that undermines the unions' class warfare, "us versus them", power-based mentality is that they frequently act in their own interest rather than that of the workers and thus reveal their true beliefs. Their anti-sweatshop activism is usually a cover for their anti-free trade, pro-minimum wage rent-seeking which works for unionized workers, but is somewhere between neutral and negative for the poor and working classes generally. Consider the recent evidence:

• Unions outsource strikers against Wal-Mart. In spite of all the outsourcing rhetoric, they seem to have found that it sometimes makes sense.

• ACORN (a non-union, living wage activism group) is opposed to unionization and minimum wage laws. Although ACORN is not a union, and is in fact anti-union within their own organization, I think this goes to show the wedge that arises between activists' stated goals and their actions when actually involved in bargaining. In their own arguments in California state court, minimum wages would adversely impact them. As summarized in the court's finding,"According to ACORN, this adverse impact will be manifested in two ways: first, ACORN will be forced to hire fewer workers; second, its workers, if paid the minimum wage, will be less empathetic with ACORN's low and moderate income constituency and will therefore be less effective advocates."

What union would ever admit to wanting lower wages to reinforce class solidarity among its workers? Not one that would likely get certified.

• Garment workers unions run sweatshops: "According to a 1999 report, the vast majority of garment shops in San Francisco, despite being mostly non-unionized, still manage to pay the city minimum of $8.15 an hour and comply with labor laws, while in New York unionized workers were making under $5.15 an hour and repeatedly subject to wage and hour violations." Yikes. This may be a case where the union locals have opportunistically used the union's reputation to exploit labor, and illustrates one reason why the national union needs to check up on the locals the same way McDonalds or any franchise needs to carefully craft franchise agreements.
  

• The 1997 UPS strike did not benefit employees as much as was claimed in the press at the time, and was probably only brought about in an effort by Teamster President Ron Carey to save himself from a widening scandal. At the time, based on the median wages that were published in the press and on their website, strike pay, and the wages lost while on strike, I calculated an 8 year payback based on 0 inflation, longer for all positive inflations. A major issue was the conversion of part-time to full-time jobs (the union wanted 10,000 jobs converted). Many part timers already had benefits, and UPS had a track record of converting part time positions to full time positions (13,000 over the 3 years prior to the settlement), so this was illusory. During the strike, UPS lost market share to FedEx and others (see note 5), slowing growth, and probably decreasing the number of full time and part time positions relative to what might have been. UPS actually offered a better retirement package before the strike, and withdrew it afterwards, but the Teamsters headquarters decided to strike without allowing the membership to vote on the package. The entire affair resulted from Teamster President Ron Carey's attempt to divert attention from an election scandal in which they were found to be illegally money-swapping with the 1996 Clinton-Gore campaign. The charges were initiated not by rabid Republicans, but by James Hoffa, who had lost the Teamster election, and confirmed by the election judges.

I am not cynical about unions per se. I believe that unions and management can work well together when the management is serious about their commitment to the customer, and when the union sees and shares their sincerity. After reading Oliver Willamson's treatment, I believe that bargaining agreements are useful to both workers and management to encourage the development of firm-specific skills that some workforces need. The representatives at the industrial and trade unions' national (ahem, excuse me) international headquarters don't share that same commitment or view: their customer is themselves, and they believe it serves their purpose to foment class warfare. For this reason, I believe Toyota and Honda are wise to avoid the UAW; Toyota in particular is able to get the best of both worlds because they have developed a reputation for dealing fairly with their employees. Local employees no longer believe such rhetoric, and that is why unions are declining today. Unfortunately, state interventions like the NLRA have locked us into the conventions popular in the 1930s.

Public employees unions? Different story. I think they are largely engaged in naked rent-seeking.