One of the things I've found over the years is that when a software project is in trouble (usually for reasons of bad code, bad design, bad architecture or shifting requirements) one of the most common tendencies in business is to throw money at it. "If we increase the budget, then maybe this will solve the problems." This is a very businessman like way of dealing with serious, endemic issues, one that can be translated into "If I increase the budget for it, then I am essentially placing the responsibility for solving this on the people I hired to do it the first time, rather than that responsibility be on me."
There are times when budget increases are warranted - initial estimates were too low based upon unknowable factors, for instance, or a change in feature set requires a change in architecture, design or previous development. These things happen, and one of the marks of a good IT manager is to know when budget requirements have changed and to plan accordingly.
However, the wholesale injection of cash into a project because the project has fallen into trouble is irresponsible at best, and potentially fatal to one's company at worst. In this sort of mode, what invariably ends up happening is that the people who tended to be the biggest bottlenecks or the most profligate spenders look upon this injection as more money to continue what they've been doing already, especially if there is comparatively little oversight on the whole process.
This process is well underway in the financial sector, where days after insurance giant AIG became a ward of the state lavish parties were being given, and one brokerage firm (I don't remember which) had started paying out bonuses to top analysts that in sum would have otherwise bankrupted the company if they hadn't just received a bailout ... though this was fortunately caught just in time (maybe).
The Dangers of Too Big To Fail
The Big Three companies (GM, Ford and Chrysler) of the automobile industry came cap in hand to Congress recently (although in many respects it was primarily GM that was really in trouble) talking about the dire change reactions that they would face if they didn't immediately get several tens of billion of dollars of funding as well - though the CEOs of each arriving in their corporate jets didn't help their cause much.
Last week's pilgrimage, this time with perhaps a modicum more humility, brought the same warnings, and GM announcing that they would go bankrupt if help didn't arrive yesterday. A deal seemed to be working its way through Congress, until a Republican initiative (apparently brought at the behest of the executives at GM in particular) also asked that any plan would require that the union employees take significant pay cuts and that GM could effectively walk away from their pension plan obligations. When the Democrats refused to consider this, the Republicans voted en masse against it and the bill failed.
A "bridge loan" seems to now be in the works from the $700 billion set aside for the banks earlier this year, so it is likely that the automotive industry will get at least some of the funds that they're requesting, but it also raises a number of significant questions about what exactly we're doing, and whether this really is a case of bad money chasing bad design.
One of the first is a concept that's been brought up here at O'Reilly a couple of times. Does too big to fail mean that failure is inevitable? The software industry is not unacquainted with the process of consolidation and agglomeration - where, rather than choosing to invest money into R&D, a large company goes and buys a startup that has already done the innovation work, expanding the large company's portfolio while at the same time removing competition from the industry. These software behemoths usually wield a huge amount of clout, but at the same time they are also weighed down by extensive legacy which limits their mobility and means that whenever a disruptive change happens in the marketplace, they are often not able to adapt quickly enough and become increasingly irrelevant.
This raises an interesting question - are companies like Microsoft too big to fail? After all, should Microsoft go into bankruptcy, it will plunge the warranties on millions (perhaps billions) of copies of Windows into doubt, it will cause the layoff of hundreds of thousands of people and will affect third party Microsoft vendors all up and down the chain. I think even people working at Microsoft would argue that this is silly - that while they personally would prefer not to see Microsoft put into that position, there are alternatives, if not necessarily ones that are completely functionally equivalent.
An editorial disclaimer here: Microsoft is not in any danger of going into bankruptcy. It is sitting in a strong cash position and will no doubt manage to weather the current economic storm just fine. The example is provided solely to illustrate a point.
One of the most important things to understand about deflationary periods (such as the one that we're in now) is that they are a lot like the inflationary boom in reverse. Companies shrink, or go out of business altogether. Conglomerates disaggregate, with corporations spinning off the less profitable businesses so that they concentrate on their core. The less profitable businesses, on the other hand, may find that with the weight of having to coordinate their activities with the rest of the conglomerate removed they are able to become more agile and responsive to market demands.
The Supply Chain Weak Links
Supply chains represent another area where the forces of deflation tend to wreak havoc. In may cases, such a supply chain typically represents satellite or vassal companies that are dependent upon the parent conglomerate while not technically being a part of that conglomerate. When the parent fails, this means that those companies in the supply chain that have signed exclusive contracts are also in trouble, because they cannot generally retool fast enough to shift towards a different manufacturer, especially if that niche is already filled by a competitor.
One of the things that emerges when this happens is that groups of supply chain companies that are all facing the same crisis find that it is a good time to talk about establishing common industry standards for these parts (something that can, and should, be helped along by government). During inflationary times, competitive pressures to build custom interfaces are high, but deflation typically leads towards establishing common interfaces, one of the key aspects of componentization. This may reduce the overall number of supply chain companies in that niche, but as an economy slows down, the benefits of maintaining twelve distinct manufacturers of starters, each with a different design, diminish pretty quickly.
Additionally, a reason innovation has proved so difficult for the Big Three is the fact that they have in fact become as locked into their supply chain supplies as the supply chain providers have become synched with their hub. Successful industry innovation will ultimately come ironically by reducing the number of near identical models, consolidating on standards and communicating with other companies to insure that such standardization can move forward in a systematic way.
This model, by the way, is hardly new - in Japan, METI (the Ministry of Economy, Trade and Industry, formerly MITI) has been responsible for providing this level of coordination for years, and Japanese automobiles have generally not suffered a bit despite the standardization brought about on many parts of a car's assembly. As with effective componentization in the software industry, componentization in the automobile industry does not mean black box components, but rather provides parts that can be adapted to fit various requirements and yet do so in a standardized manner.
Does Protecting Jobs Protect Workers?
On the other hand, the recent setback in the Senate point to a much more uncomfortable aspect of this Schumpeterian creative destruction - if a giant company like GM goes bankrupt, lots of workers will lose their jobs (or at a minimum their pensions), and/or lots of investors will lose their dividends or even their capital investments.
Labor is almost invariably the most expensive part of any production process. Ideally, from a management standpoint, you'd like a business that was completely automated - machines can work 24/7, don't require time for lunch or breaks, and essentially represent a fixed cost - you don't have to pay a machine more the more training and experience he or she has accrued. You can replace a machine without having to worry about unemployment insurance and so forth.
Unfortunately for the owners of most companies, even the best automated system still requires people to install and inspect the machines, to design the cars, establish tests, and then handle the administration of those people. You need people to sell the cars, to set up advertising campaigns and handle financing.
Moreover, there's the great paradox of capitalism. Someone has to buy those cars that you manufacture, and to do that, they have to have a steady, reliable paycheck. As the buying power of people decline (because wages are in decline or because companies attempt to get by with fewer and fewer people) then fewer people have money to purchase your cars, even on credit. In general, the better paid people are overall, the more likely you are to sell your cars, yet the general trend in management has been to reduce salaries.
This is where unions come into the equation, as well as what will likely be collectivist organizations that don't necessarily fall into the union model. Unions establish for their members a base salary and benefits package, establish rules for overtime work and provide protection to their members. Unions have generally been beaten back dramatically in the last thirty years by successive waves of "pro-business" legislation and business-sympathetic political leaders on both sides of the aisle.
There is a problem with unions, however. Unions are, by nature, protectivist organizations - they tend to protect the job position, rather than the individual, even when the position is no longer relevant. This can play hob with a company's ability to be agile, and in some cases, such as a number of Hollywood unions, simple tasks can become inordinately complex because the union has won the right that only a particular union member can do parts of that task, even if it is something as simple as moving equipment around or replacing light bulbs.
Overall, these are relatively rare phenomena, however, and typically occur as a way of making a point when an organization's management is generally trying to reduce the influence (and by extension, the labor cost) of the union.
Ultimately, though the question that is being addressed in the bailouts is whether the goal of bailouts is to better support the American worker (similarly issues are being faced in other countries) or to attempt to make automobiles, trucks and utility vehicles more competitive both at home and abroad. My argument is that the current approach in fact does neither, but nor does the idiocy of attempting to shaft workers in the name of protecting an existing industry.
Breaking Up Is Hard to Do (But Maybe Necessary)
I'd contend, in fact, that it is time for General Motors to be split up into separate automobile companies, for these companies to agree upon the aforementioned discussions on componentization, and that these companies individually be allowed to succeed or fail. Some will and should - one of the dangers of conglomerates is that it makes for non-competitive products to remain around for years because they are subsidized by other divisions. During said breakup, the pension plans would be broken down into more manageable pieces, and possibly even broken off into its own company in much the same way that GMAC finance exists as a quasi-distinct entity today. (I'll leave health care for its own discussion later).
The result of this would be a constellation of much smaller corporations, some emerging as design firms, others becoming supply chain providers, still others providing accounting, human resources, or pension management. Government's role in that case would be four-fold:
First, provide the funds to help in the disaggregation of GM - while no doubt the senior executive management of GM would likely balk (no one likes having power taken away from them) the reality is that if GM is too large to fail, then it's too large, period. In the end, the stockholders benefit (divestments of companies typically create a number of smaller companies that, in the aggregate, have higher values over time), the unions benefit (it increases the total number of jobs and opens up opportunities that wouldn't have existed elsewhere), consumers benefit (the smaller companies, no longer tied into a larger corporate strategy, are considerably freer to experiment with new designs and technologies), and the government benefits by increasing competition in the marketplace by making it easier for those companies that aren't competitive to fail gracefully without bringing down the entire economy, and possibly by gaining a stake in shaping these new technologies.
Second, work to formally transform standards in automobile design, production and environmental integrity so that they can work across the industry, rather than just one coalition. Overstandardization can occasionally dampen innovation, but this usually comes at the end of period of consolidation, rather than the beginning (though by and large most companies will tend to argue that such standardization is bad for their respective industry at any time ... especially if they happen to be the market leader).
Third, establish goals that would provide a general guidance into the needs that the government (and the country) is seeking to accomplish. Turning GM into a number of mini-GMs really serves no purpose if each of these simply continues to perpetuate the mistakes that were made under the larger organization. In order for business and government to work together, government has to be the one establishing the mandates and direction. If a business chooses not to follow that direction, that is the decision of the management team and board of directors, but clear and demonstrable goals that will in turn become a sizeable market if the companies in question are able to achieve these respectivegoals.
Finally, demand accountability - a business incorporation license is not a license to commit fraud. Place financial barriers on CEOs and Boards of Directors that reduce their ability to attempt market manipulation of a company's stocks rather than substantive innovation. Make C-level and board member salaries contingent upon performance, and determined not by the Boards of Directors but by the shareholders themselves. Go to regular (quarterly at a minimum) electronic submissions of a company's books in XBRL or related formats in order to force some transparency into the respective industries.
Note that while this discussion has been centered upon the automotive industry, as a model it is as applicable to other domains as well, especially those that have become so centralized around one to three primary players that innovation (and hence adaptability to market conditions) has been reduced to near nothing. This ultimately gets back to competitiveness - make your company more agile and responsive to the market, establish clear but ambitious goals that don't necessarily translate into profitability this quarter but that can make such companies competitive longer term, add disincentives in the equity markets that will keep companies from growing into oligarchic behemoths that completely dominate the industry.
Government absolutely must play a role in dealing with companies that are too big to fail ... it must prevent them from reaching that point. Until that happens, real reform in business will be slow and problematic.