Sears: A Libertarian Case-study

Great Debates
141

Every large enterprise has a problem with information flow. Very large corporations like Sears can become non-competitive if they rely too much on top-down management, for the simple reason that the managers in the head office lack the local knowledge that managers in the field may have.

Every company struggles with the need to compartmentalize and delegate authority, while still maintaining a coherent business model and keeping everyone aligned on the same goals.

The only way I see ‘libertarianism’ coming into this at all is that information transmission is generally seen to be a big problem for centralized governments, and libertarians believe that by maximizing freedom you can ensure maximum latitude for people with local knowledge to make decisions based on that knowledge.

If you want examples opposite of Sears where decentralization worked, I can give you two quick ones:

When IBM first tried to make the PC, it suffered from a sclerotic, top-down management structure. If found itself unable to innovate, and there were too many vested interests in upper management who saw the PC as a threat or competitor for corporate resources. The answer was to spin off a PC division as an autonomous unit and free it from the constraints of a management culture that didn’t understand PCs and couldn’t operate at the speed the new PC industry required. This worked very well.

Jack Welch’s restructuring of GE is another example of the success of this kind of decentralization. When Welch joined GE, the corporate structure looked a lot like a government - there were 46 different divisions, each of which had large reporting requirements. All of this information flowed up to the head office to ‘strategic planning groups’ that made decisions about corporate direction and then issued directives back down to the various divisions. The information flowed through multiple levels: department managers, sector managers, subsector managers, unit managers, supervisors. The result was a company stuck in an old-world paradigm, unable to innovate quickly or shift to meet market demands.

Welch attacked the bureaucracy. He fired half the strategic planners, and gave authority back to the the individual business units. He got rid of the complicated plans and stragetic initiatives and issued a simple directive: Either your business will be #1 or #2 in its market, or you will be sold or shut down. So he gave his managers leeway, but he also held a pretty big carrot and stick over their heads.

He followed through on his threats, and got rid of all the chaff. The bureacracy was decimated in favor of more local autonomy. I worked for GE in the Welch years, and he used to send out E-mails saying things like, “If you spot anyone trying to build a bureaucracy in this company, let me know and I will fire them.” Woe to the manager who tried to prevent his people from taking concerns to upper management or who tried to control the flow of information so it had to go through him.

Another thing Welch did very successfully was to foster compeititon between business units. For example, under the old GE structure, if one division needed a product that another division made, they were forced to use that division’s product. Under Welch, we were free to buy the best product on the market, whether or not we made it. So even though GE made a certain type of machine part, if another division needed that part but felt a competitor’s version was better, he could just buy from the competitor with no ramifications from management.

At first, this might seem crazy, but in fact what happened before that within GE was a type of ‘crony capitalism’. If a maker of a widget in GE knew that he had a guaranteed market of X thousand widgets by selling to other GE divisions, he could get sloppy. He had a monopoly on GE business, so why bother with ultimate quality? Of course, there were lots of ‘directives’ to make good quality stuff, but nothing motivates like building something and finding out no one will buy it. In the meantime, the businesses that were forced to use non-competitive materials from other GE businesses found themselves at a competitive disadvantage. Overall product quality suffered, and the company got lazy.

As a governmental example of this, consider what’s happened to NASA. Its programs are always bloated by requirements that come from cronyism. The Apollo program had to have its launch facilities in Florida but its mission control in Texas - not because this was the best way to go into space, but because certain Senators and a certain Vice President had to be mollified. The replacement for the Space Shuttle is a boondoggle because of the political requirement that the new system be built from Shuttle parts, simply to prevent laying off work forces in key political states. In the meantime, free-market competitors like SpaceX who have no such constraints are doing it better and faster. This was the kind of thing that was happening in GE and was preventing it from improving quality and innovating.

All of Welch’s reforms were along ‘conservative’ or ‘libertarian’ lines. Rather than creating new grand central plans and strategies, he axed the central planning. Rather than building internal regulatory agencies for ensuring that products were built correctly, he simply introduced market forces. Rather than forcing all information to flow up to the central authority for analysis and subsequent central command, he streamlined the company and gave the individual units more autonomy.

Another ‘libertarian’ reform Welch undertook was to allow bottom-level employees much more say in how things should be run, and to install a new pay system and benefits package that rewarded individual initiative. We held ‘work out’ sessions where the managers pulled together site staff and allowed them to voice ideas and complaints, which were taken very seriously. Rewards programs were put into place for employee ideas, and managers were threatened with termination if they tried to hide employee criticism. GE developed a strong ombudsman culture were anyone could cut through the bureaucracy for any reason and no retaliation would ever happen.

The result of all this was that by the time Welch left GE had been transformed into a modern technology company, its revenues more than tripled, while the table of organization was flatter and the amount of central management was much smaller.

Does this prove that libertarianism is correct for governments? No, because enterprises aren’t governments. However, it does show that when corporations get very large they can suffer from some of the same problems that governments share - bureaucratization, cronyism, inability to make decisions quickly, lack of information on the part of decision makers, etc. These are not trivial problems, and the answer to them is not to add in layers of new bureaucracy or hire more central planners, which seems to be the strategy Democrats generally want to take.

142

I can offer another example of where central planning has been replaced by bottom-up decision-making very effectively: The new trend towards Agile and Lean practices.

The old waterfall programming paradigm is much like a central plan: You have a whole hierarchy of analysts, architects, marketing people, and other central planners. They go out and gather requirements for a new software project. Then they collect all these requirements and come up with a feature set. The architects take that and figure out how the software will be built and they create a series of documents to communicate their architecture to the programmers below. This process could take months or years, and hundreds or thousands of pages of technical documents, business rules and the like would be handed down to team leaders who would then break it all apart into chunks for the developers to build.

The problem with this is the same one that central planners always run into - complex central plans never survive their run-in with the real world. Ask the architects of Obamacare, or the designers who tried to digitize the post office and failed. So developers start building this behemoth of a project, and run into issues the architects never thought of, or customers would not like the grand design being foist on them, or flaws in the original design would be uncovered. Or perhaps assumptions about the productivity of the development teams weren’t correct because the guys who came up with the plan didn’t have a good understanding of the skill sets of the workforce or other challenges they faced.

In addition, these large waterfall projects did not scale. Once you get past a team of a certain size, trying to manage the flow of all the various pieces to deliver something customers wanted on schedule and on budget becomes almost impossible. There’s just too much complexity.

Introducing change into this process is extremely expensive. A design flaw discovered when QA testing the finished product could totally derail a project. An undiscovered flaw that requires extra work could derail the project plan and cause the schedule to explode. Trying to find work for all the developers such that they all finished their pieces when needed becomes almost impossible. On a huge project you need an entire team of project managers just to try to keep everything moving forward in sync. This adds costs, and it also inhibits necessary change. Bad ideas or substandard designs get built anyway, simply because it becomes too expensive to change it.

The result was that all the grand plans would often become obsolete the minute they were finished, and software projects invariably came in grossly over budget and behind schedule. In the worst case scenarios, the top level team would completely lose control of the project and the whole thing would collapse under its own weight.

The computer industry’s fix for this is to move to Agile, where design is iterated from inputs that come from the bottom up. Rather than build something that people in head office think customers want based on information collected by analysts, it’s better to just build something and let customers try it, then find out what works and what doesn’t and modify accordingly.

Rather than creating a central plan for a product that will never change and therefore has to be figured out in minute detail, you focus instead on writing software with refactoring in mind so that it can be easily changed once real-world requirements are discovered, and as the requirements change with a changing marketplace. Rather than huge, centrally managed multi-year projects, you minimize risk and improve quality by running smaller, faster projects on short development cycles, then repeat as lessons are learned from the field.

The parallel with what the government is doing is pretty obvious. Take Obamacare. In this case, a bunch of smart central planners try to take into account all the nuances and requirements of a massive health care system, come up with a ‘comprehensive’ plan, and then issue rules to force everyone to implement the plan as written. And when the plan starts to come apart when it contacts the real world, they’re left with a mess.

In Canada, on the other hand, we operate our health care system much like GE operates as a company. There is no detailed central management. There’s no federal agency that has the responsibility of figuring out what the health care system for New Brunswick or Alberta should look like. The Canada Health Act is essentially a six page mission statement instead of thousands of pages of regulations and directives. Each province is left to implement the details of their health care system, and some provinces further break down their health care authority into smaller, local authorities.

This isn’t directly analagous to libertarianism vs statism - After all, Canada’s system is by no means libertarian. But it does highlight a lesson Democrats should learn: If you want effective government, perhaps centralized control of minutae by Washington isn’t a good way to go about it. Perhaps rather than having a huge Department of Education setting national rules, it might be better to delegate authority to the states and school districts and just issue general guidelines from Washington. Instead of additional departments and bureaus and regulatory bodies, it might be better to figure out how to empower states or local authorities. Rather than gigantic omnibus regulatory packages, it might be better to take a more incremental approach to solving problems, then iterate quickly as problems surface.

143

Sam - vis a vis IBM and GE, I think you’ve put your finger on 2 exceptions that prove the rule. IBM found itself with 2 business units that had evolved in different directions and had a declining amount of synergy. So they made them more stand-alone. Then they sold the PC business. That’s exactly what you would expect would happen, if you were a believer in Coase, Williamson et al.

In contrast, I would say that there are definite synergies between Sears’ IT department and its Craftsmen division. Arms length transactions make a great deal less sense.

GE is a conglomerate, which is basically the definition of a group of units with weak synergies. There are 2 sound strategies in that case. The first is to break up the conglomerate, which is what was done most of the corporate monstrosities formed during the 1960s and 1970s. (Conglomerates can actually make sense in the 3rd world, but that’s another matter.) The 2nd is to… read Jack Welsh. Seriously, there are very few who can make that business model work, but Neutron Jack figured it out. Part of his approach is to sell off units that are not #1 or #2 in their field. That’s fine, but it’s not a strategy that can be replicated a lot: you would expect only 1-3 such companies to exist.

So again, Sears (and most all firms) is different than GE. The central problem that Eddie made was not considering greater unit autonomy. The problem is that he never seriously considered problems of cannibalism that were blindingly obvious to anybody with experience with department stores. Heck most MBAs could work this out.

A better example might be Neimann Marcus. Last time I traipsed through there (years ago), I noticed that it was set up as a collection of boutiques, each associated with a brand name (Chanel, etc.) That’s the sort of process you could atomize, though I trust that NM’s CEO didn’t appoint a board of directors for each and every unit, as Lambert did.

144

I don’t disagree with any of that, other than that you seem to be implying that these examples are rare exceptions to a norm of strong top down management.

I believe that is true at a certain size, but as a company gets very large the need to decentralize grows, as the complexity just becomes too great. The answer sometimes is divestiture, and sometimes it’s to a more decentralized management model.

How many companies solved their growth problems successfully by adding layers of new bureaucracy and by centralizing decision-making?

145

Post #70!

:smiley:

146

You’d think it’d be so self-evident though, what with the increased freedom and rainbows and sunshine for everyone. Why hasn’t it been popularly tried? Is the problem that there are obvious errors with libertarianism or simply its bad at marketing itself?

That’s besides the issue. How many successful businesses run on the libertarian model? Apparently none, because people need to be led and governed

I’ll explain. The Sears model of failure used the libertarian hands-off approach, allowing different departments of the same corporation to compete with each other, ensuring their failures because there was no top-down leadership to organize the resources of selfish independent departments with competing interests. Thus the screwdrivers and lingerie debacle

A more socialist business model accords ad and shelf space to weaker departments to prop them up, realizing that part of the strength of a department store brand like Sears is its all-in-one diversity of products. So what if, for example, electronics are making the most money? They are doing fine on their own, so there is no need to allocate resources to them. In fact, take some of their money to help expand the children’s clothing line that is not doing so well. Looking at it in a macro fashion, such “taking from the rich and giving to the poor” hones the strength of the Sears brand. There’s plenty of places that I can get just electronics or just kids clothes, but I can go to Sears to get both for a cheap price.

Rewards can still be doled out for high performers, but it is better for the brand as a whole, in this case substituting for The Government, to be strong across the board than to excel in certain areas.

Libertarians don’t get this. They see themselves as a man on an island; either it helps them or doesn’t, and if it doesn’t, then screw it. But your home’s worth is a factor of your neighborhood, the resources it has access to, and your neighbors despite how much libertarians want to ignore that. If you pay for taxes that keeps your neighbors’ kids in school, then it means there will be less crime, less poverty, and in turn that helps you. But if you take the selfish libertarian route and say you’re not a parent so why should you pay taxes for school, then you will see how the libertarian model of selfishness fails in all regards, business and government, and that it will never ever work if you have more than a handful of people that is using it

Anyways, that’s just one example. Its not the only one, nor am I saying things can’t work with some libertarian influences. But as a whole, libertarianism doesn’t work because its rigidly enamored of itself and refuses to compromise. Taxes are a good thing, regulations are a good thing, public services are a good thing, and having a fairly large government for the over 300 million people we have in this country is a good thing. Only when it because too much are each of these things bad, but the answer is not to have the minimum possible

147

Libertarianism for all its faults does not advocate people be ungoverned. You seem to think libertarianism is the same an anarchism.

I actually find it shocking liberals want to use libertarianism as such a boogeyman though, it’s an extremely marginalized political philosophy not in the mainstream in either party. It’s also strange that you guys are equating any form of decentralized decision making with libertarianism. You could have an absolute dictatorship that was highly decentralized, you’d just have to have a top level dictator who wanted to delegate authority, but he’d have ultimate veto power and could recall his under-czars or whatever at will.

Decentralization doesn’t strictly have anything to do with libertarianism. Eddie Lampert, if he was really trying to structure his company on libertarian ideals, did something akin to training a team of draft horses to fight against one another which is very counterproductive. Libertarianism doesn’t advocate making the internal units of a business compete with one another, but instead would say businesses should be free to use whatever sort of organizational form they wish.

148

You’d think women’s suffrage, or free trade, or separation of powers, or marriage equality, would have been self-evident too. And yet, they clearly were not.

If you’ll excuse me, I have to go scold marriage-equality advocates, as their ideas must have obvious errors or they must be bad at marketing themselves.

The actual statement you made is besides the issue? Odd.

Also, libertarianism includes leadership and government (except for the anarcho-capitalist loons).

Sears cannot stand in for the government in any sort of intelligent allegory. Firms and governments have wildly different goals and structures.

The divisions of Sears are part of it in a way that independent firms are not part of the government. Sears takes all of the revenue generated by all its divisions, regardless of how the org chart is set up. It’s a closed system. Nothing is actually independent, which is why it was a stupid idea, like having both your feet compete with one another. In the actual market, where firms are independent, competition works marvelously well.

Your socialist business model has the same problem and is also deeply flawed as an allegory. If a department keeps using more resources than it generates, then Sears will shutter that division lest it drag them under (or attempt a radical overhaul). If a citizen uses fewer resources than he generates, the socialist government will not kill him, it will keep giving him what he needs. The government is not competing with other governments that force it to keep things lean, it has no owners to satisfy with revenue growth. The actual market is an open system. And in the actual market, cooperation instead of competition produces shortages, shoddy goods, and corruption.

149

I know, as I said I’ve been through it. The reason that the stuff you mention happens is that it is very hard to manage to metrics about corporate profits, and much easier to manage to departmental profits. “The good of the company” doesn’t get you a bonus unless one of the metrics is the good of the company, which it clearly was not here. And it is not just competition or screwing the other guy. The IT example I mentioned does not directly hurt any other department, but could indirectly hurt the company by preventing the sharing of information.

150

First. companies I’ve seen who use a business unit model give goals and funding to the units, and expect local management to meet those goals, subject to some big rules. The sort of conflicts that arose at Sears should have been resolved one level up.

IBM’s issue was less top down management than the usual blindness to something new coming when you are successful. Microsoft clearly suffers from the same problem. My adviser organized a Problems of the '80s conference in 1978 on the future of computing.. I reviewed it recently. People like Adam Osborne and Portia Isaacson were amazingly perceptive on the impact of small computers. People from IBM universally saw the future as mainframes now and forever.
Remember also that while IBM invented the PC and IBM’s clout validated the PC, the way the division did it quickly meant they lost control of the PC, so today they both have no PC business and a much less dominant hardware business. Good for the industry, sure, good for IBM maybe not. They also do quite well tying services sales to hardware sales.
I’m sure you’ve heard customers complain when multiple divisions of a company make multiple sales calls selling somewhat competing products. That happens when you have too little top down control.

151

Saying things like that, appealing to the “its always been this way for others” doesn’t strike me as a very effective way of articulating why, right now, no country has ever tried it. Freedom hasn’t been a dirty word since the world came together and most of western civilization threw off the yolk of monarchy. So its not enough to simply claim that libertarianism falls prey to the same forces holding back women’s suffrage, you’d have to actually come up with a reason why it hasn’t been tried in one of the many states around the world who both value freedom and prosperity enough to give it a shot. There’s another debate thread going on here right now stemming from Michael Lind’s essay why libertarianism doesn’t work. If you could tell me that without resorting to unnamed factors and history holding you back, I’d appreciate it

Not odd, because I was responding to another who brought it up in the first place

You’re dodging the issue. Clearly Lambert is a capable leader who has some skills to both become a billionaire and to get to the top of Sears corporate. The question you’re avoiding is: why has his libertarian-based business philosophy lead to Sears’ decline? I suppose you could answer by responding to the examples quoted in the OP such as “How has libertarianism failed when unlike items such as lingeries and screwdrivers are promoted side by side?” and “Describe the negatives of allowing richer departments who outbid poorer ones for ad space”. I want specific examples, I’m uninterested in a regurgitation of vague libertarian notions of freedom

So you admit that the libertarian-based business model for Sears failed because it was libertarian?

It works well only if you’re not a user of one of the failed companies or services. Many of the poor, which libertarianism would create in abundance, would not be better served when markets compete with little regulation or a government-funded cheaper option

Your mistake, which was completely unprompted by me mind you, is thinking that such a socialistic business model automatically creates departments using more resources than it generates. You are assuming facts to fit into your worldview instead of looking at what reality is. In fact, I made no mention that the children’s clothing department was losing money, only that it was weaker than the electronics. You chose to believe that weaker automatically means a drain, that’s your mistake. My socialistic business model would not create a drag on the business at all. Weaker departments are still making money, just not as much as the strongest ones. And there’s no need to shutter the weaker ones to focus on the stronger ones, they all help each other. Its simply that the weaker ones may need more resources to keep it going and the stronger ones do not

Wow, do you even realize what you’re implying? You’re implying that the bolded is a BAD thing, whereas in your libertarian utopia, that government WILL kill him and that its a good thing they do

152

I already did: it’s not something that a dictator would ever install by force, and in democracies, not enough people want it. They are either ignorant of it (which includes ignorance of the nuances between, say, anarcho-capitalism and moderate libertarianism), or don’t think it will benefit them more than what they vote for now. It’s not some big mystery.

I already wrote this too: it’s a wildly unsuitable business model for retail, whether it was inspired by Ayn Rand, Marx, or a fever dream.

Yes. In this specific context, it was a bad idea. That doesn’t mean it is a bad idea in every context. Having two captains alternate in choosing players is a great model in the context of pick-up basketball. It’s a terrible one for, say, Fortune 500 companies to use to fill their staffing requirements.

It would be highly unusual for a political philosophy to double as a retail business model.

Irrelevant. The thread is about Sears’ practices proving libertarianism to be an unworkable political theory. It cannot prove that because the internal structure of a business is not analogous to a nation.

Your socialist business model is unlike actual socialism, and just as useless for evaluating it as Sears’ is for evaluating libertarianism. Actual socialism must account for the un-and-underproductive, you cannot just stipulate that none exist (unless you do kill them, which is hardly unheard of as socialist policy).

No, I’m not. I’m pointing out that one of the many ways a business is unlike a nation, and that strategies for one don’t translate to the other, is that nations don’t have the option of eliminating underperforming divisions and workers, unless it is lead by murderous tyrants.

I have no libertarian utopia.

153

Oh there certainly are very real tradeoffs between coordination and sclerotic red tape. Just as there are tradeoffs between beneficial economies of scale and the initiative of smaller business units.

But it seems to me that Lambert is your garden variety bad CEO. He’s a numbers guy, so he skimps on things outside of his comfort zone, like understanding the business, face to face meetings, or interpersonal relationships. (Jack Welsh incidentally was a master at all of those AFAIK.) It’s one thing to assign metrics to every business unit. It’s another to mandate that each have a freaking board of directors. Because he could. Lampert combined the worst aspects of red tape, decentralization and executive clueless hubris. For an example of cluelesness, see his adventures with the in-company social media utility Pebble: Sears CEO Internal Facebook - Business Insider *

Again, if you want to go towards full bore decentralization, you should divest. If you don’t divest, you are in the land of tradeoffs, something that was apparently alien to those with market fundamentalist dispositions.

  • Control freakishness:
154

First of all, Human Action, thank you sincerely for acknowledging that in this specific context, libertarian philosophy was a bad idea. Too many libertarians hold their views to be the best and only way to do anything, whether it be business, government, or picking toppings on a pizza

There are obviously many flourishing democracies. Are you saying that ignorance of the nuances or its benefits is the only thing holding it back?

Personally, I thought it was about proving it was unworkable in a business setting. Someone else brought in politics

Then I’ll agree that, given that there are some drains on company coffers for certain departments, socialism benefits diminish in those instances

Why did you point out specifically that such a socialistic government will not kill the underperforming citizen? All governments short of tyrannical ones, do that. Your critique is not unique to socialism

155

Well, they are simply wrong. Absolutism is nearly always so.

No. When I wrote “or don’t think it will benefit them more than what they vote for now”, that wasn’t meant to imply that they were incorrect to think that. I claim no monopoly on what’s true or what’s best. I happen to think that a more libertarian (as opposed to radically libertarian) government would result in a net increase in the quality of life in the United States. By “net increase”, I mean life might get 15-20% better in the aggregrate; some things would get better, others worse. There are no utopias to be had.

But, I could well be wrong. Others have considered the same ideas and feel differently. Perhaps they are right to do so. Unless and until enough people change their views and start voting for libertarian candidates, we won’t know. I’m not in the habit of assuming that the current policies are automatically the best ones, and I bet you aren’t either. Consider Social Security. We had a flourishing democracy for decades before it was enacted in 1935. Were all the voters before then ignorant of the nuances of such a program, ignorant of the benefits, or was there some other reason it took so long?

The OP did.

Agreed, though, again, not comparable to a socialist state.

To highlight the folly of using business practices to credit or discredit a political theory. Businesses don’t behave as states do, and the same policies that serve one would be disasterous for the other. For example, a business can eliminate a burdensome department and no one will bat an eye, it’s certainly not considered wrong beyond the general misfortune of laying off workers. If a state were to eliminate burdensome citizens, using the same reasoning the business used, it’d be a gruesome crime.

If using business policies to prove or disprove the validity of a political policy wasn’t what you intended, then disregard.

156

The thing is, all political ideologies are inherently economic models, because a country’s economy is vitally important in it being able to maintain its way of life. It doesn’t mean the GDP needs to be the best, but it at least has to be above expenditures. A government has to be profitable.

So if libertarianism fails for a business, it does mean that, at least, in some cases, it would fail for a country. Then the question is, defining what is so different about a country than a business. The best I can come up with is that our country is a democracy, while Sears absolutely was not. Therefore people weren’t really free enough to go after their own rational interests.

Would you guys agree that you can’t have a libertarian dictatorship? That you can’t have a king from on high saying that you must be libertarian–you must become libertarian by choice?

157

In that case, the only political ideology that wouldn’t fail for a country is a dictatorship, as that is how businesses are structured. Depressing, if true, but looking around the world shows that it’s not.

You could, but such an enlightened despot, a dictator without dictates, would be a rare bird indeed. The closest thing to that in history would probably be post-war Hong Kong under British rule, when men like John J. Cowperthwaite pared back government control over the citizens and the economy.

158

Bhutan’s king is trying to push his country into being a democracy

159

Neat, and good for him, but that means Bhutan would cease to be a dictatorship, and thus couldn’t be a libertarian dictatorship.

160

Ah, was reading on the phone, and mistook your point.