There are effective ways to sharply improve the outcomes from an organizational process or processes. The free market ideologues would lead you to believe that the only way is to let private business operate completely free from any and all government interference, and that somehow the magic of the market will provide great products and services at the lowest possible prices. There are elements of truth to that, but the idea is more flawed than accurate. Let’s look at how a process can be made to provide sharply improved outcomes over those it provided earlier.
In order to improve a process, a decision-making person has to have feedback on the quality of the results of what was done earlier, and that must determine which efforts succeeded and which failed. Then they must learn the “best practices” which can increase the number of successes and reduce the number of failures. This requires a thorough understanding of the process that provides the outcomes and very careful measurements of the most important aspects of that process. The very striking example Dr. Atul Gawande presents in his book "Better: A Surgeon's Notes on Performance" (Metropolitan Books, 2007) is the way the military has measured emergency medical care for injured soldiers and improved that care so that deaths have been reduced from 25% of causalities down to 10%, not by breakthrough new medical methods as just by applying best practices to the points in the process with the highest death rate.
The key is to isolate a process, or even a segment of a process, and then measure everything that is happening in that process and also to measure and rank the outcome of the process according to its degree of success or failure. Then focus on what elements of the process contributed to the failed outcomes and determine how those elements can be changed. The process must be thoroughly understood. I am not sure whether Dr. Gawande realizes that this is an application to battlefield medicine of the ideas called “Best Practices” in business quality control, but it is.
One very important point to keep in mind is that after a change in the process outputs will often degrade for a while until the newer process becomes established. This temporary period of degradation must be accepted if the overall process to become better in the long term.
Another very important point is to consider how the process is isolated. There has to be an organization that is inside those boundaries and which encompasses the problem that is to be improved. Those boundaries are of critical importance. People outside those boundaries must not be allowed to interfere with the decisions taken by the qualified decision-maker who is working to improve the process.
Andro Linklater reported on an interesting historical experiment in allowing individuals set up relatively autonomous organizations, measure their success and improve them in his book "The Fabric of America: How Our Borders and Boundaries Shaped the Country and Forged Our National Identity" (Walker, 2007.) He contends that it was not just the frontier that made America such a successful nation. It was also the fact that Americans applied the latest survey techniques to the older common law ideas of private property as Americans moved west into new frontier areas. By use of the latest surveying technology and by having the government record who owed each parcel of land American ranches and farms became independent little organizations, each managed by its owner.
These individuals then had their own personal property with which to make economic decisions. Output could be measured in terms of amount of a crop or number of cattle, and using markets to put prices on those outputs, prices could be used to compare relative outputs between farms and ranches. This allowed each farmer or rancher a sphere of independence in which no one else could step in and make decisions for him. The most successful farmers and ranchers were those who learned and adopted the best practices.
This isn’t really a new idea. It was, in fact, the reason for the state agricultural extension services and the agricultural schools in America. What it was, however, was unique in world history up to that time. The very economic success of this process is largely the basis of the free market ideologists in politics today.
There are a number of problems with this model, however. First, and most important, markets only provide a limited measure of economic success or failure. Productivity matters – a lot – but so does quality of life for people in and around a business. The sawmill up river from the town I grew up in threw its chemicals into the river for over half a century. That lowered their costs, increased their profit and destroyed the fishing (among other noxious results) for nearly 100 miles downriver. Market-based economic decisions could not, by themselves, measure all of the costs of the lumber produced because there is no connection between the market transactions for land, labor and capital inputs and the resulting corporate revenue accept the (self-serving) assertions and estimates of business management in their financial statements. Government has to provide additional measurements for such otherwise unmeasured things as environmental dumping. This always creates major political issues around what should be measured and published as measures of the real costs of input and output.
This becomes an even greater problem for organizations that are not economic in nature. Schools are an example. In spite of all efforts we have yet to develop comprehensive effective and informative measures of the important characteristics of the most important inputs to a school, especially the characteristics of students and teachers. On top of that, there are no markets that create a semblance of comparable measures of student outcomes. Grades at best record the opinions of teachers about short-term progress, or the outcome of very limited written tests. Nor do we understand the educational process well enough to describe it, let alone comprehensively measure the most important elements in it. The result is that every attempt to apply objective measurable standards to inputs and outputs of schools is met with political cries of outrage by those who don’t like the results of those measures.
Justice Roberts recently solved that problem very neatly in the question of determining the effects of Racism in schools. His decision was that measuring race is, itself, Racist, so it was forbidden to measure race or racial outcomes of schools. The American problem of Racism is one that he and four of his colleagues do not want solved, so they have the power to prevent measurement of the problem. They used that power.
Another problem is that as the American economy has become more interlocked both internally and internationally, the level of problems to be resolved has grown larger. The decision domain of the solution to a problem must be as large as the problem itself for any solution to work. Larger organizations exist to control larger (mostly economic) processes.
Also, as economic processes get larger, they begin to have to differentiate between those controlled by markets at each stage of production or by common ownership, which allocates resources administratively, rather than by free markets. A steel company that owns the iron and coal mines, the foundries, and the steel fabricators will rarely resort to using markets between the mines and foundries and again between the foundries and fabricators because the information collection costs of markets is very expensive. The overall cost of the final product can be greatly lowered by using administrative means to determine how much coal and iron is mined and shipped to foundries and how much steel is shipped to fabricators. However, the increase in efficiency is attained at the expense of flexibility, so that such a large organization cannot be innovative the same way smaller ones can be. The strategic requirement for such a large integrated company to compete on the basis of cost per unit means that major changes in the product cannot be made without losing market share during the period of the time in which the changeover is being made. (See Oliver E. Williamson and in particular his book “Markets and hierarchies : analysis and antitrust implications : a study in the economics of internal organization”)
The limited decision domain around the problem must be large enough to encompass the problem. If the problem is that trains do not leave a given station on time and the decision domain is established at the level of the station master, he can keep the trains from sitting at the station too long but he has no control over whether they arrive at the station late. A decision maker cannot completely deal with problems from outside his area of control. That is a major reason why economic organizations get larger. But the managers who work closer to the economic processes make only those decisions delegated to them from higher management.
The American agricultural economic experiment breaks down as businesses get larger and decisions are centralized further away from the individuals who physically work with the core productive processes. Government has always existed to solve society-wide problems, but as economic processes get larger, the organizations that control them get larger. But like government, top managers in larger organizations make their decisions primarily on the basis of political power of the decision makers, since they cannot have direct experience with all the processes they manage. Direct experience is possible only in small organizations.
The American agricultural economic experiment applied historically to decisions made in the limited domain of relatively small family farms, which is why the idea of establishing boundaries and measuring processes and outcomes carefully within those domains is important. Problems that are too large to be dealt with effectively within a given set of borders require that larger borders be drawn. But what happens when the decision is too large for any individual to handle by himself. The decision-makers in such very large and complex organizations require staffs of specialists to assist in decisions in their respective areas.
Decisions in such an environment cannot be made by a single all wise individual because individuals are limited in how much each can rationally decide. That is why the decisions made at the central core of every large organization will always be political. That is, the decision-maker with the greatest level of power will choose which of the alternatives will be implemented. But since the decision is based on political power, any negative outcome of that decision will cause the decision maker to lose power. Any measurement of negative results will either not be measured or the measurements will be spun to provide good results. It is impossible to improve processes when the outcome measurements are not honest, but no political decision maker dares honestly report negative results. Improved outcomes result only from determining what processes cause negative results, and that is impossible for a political decision maker to permit.
It is because of this process that rational decisions in an organization must always be made at the direction of a politically powerful individual.
Problems that are too large to be solved within any given organization normally have to be excluded by the decision maker instead of solved. As an example, when Racism is a major problem in society that problem will cross over any decision borders that are defined for groups smaller than a state or nation, and will often cause problems within the smaller decision domains. Racism, being a social problem, cannot be solved in each of the set of limited domains. The minority Race can, however, be excluded and this will often lead to better outcomes for those limited decision domains because of reduced conflict within each domain. As the American experiment with segregation proved, such problems grow larger in society if they are not solved. The social conflicts and the waste of productive resources limit both the economy and society, so a larger entity with more than just economic power will have to establish Race-neutral best practices and enforce them on the more limited domains. Only government can make such social changes, and the otherwise economically independent (but smaller) business groups have to be changed whether they like it or not.
Like any change in the process intended to improve the outcomes, such an edict will have the effect for a while of reducing preferred outcomes until it is fully established. Only after it is fully accepted will the better outcomes become apparent.
The advocates of totally free market solutions will not allow each employee of an organization to make their own independent decisions, but they have to let trained individuals work as professionals. A nation or society must also have the same flexibility with respect to the economic organizations within it. Social problems have to be solved just as well as economic ones do.
No comments:
Post a Comment