I've read with interest the recent diaries on trade deals, foreign suppliers, and food safety. These are hot issues. However, many diarists and commenters don't really know the ins and outs of the global supply system and outsourcing. They know the problems, but not the mechanics - and details can matter.
Since it's been my academic field for 20 years, I'm starting a diary series to explain some of the mechanics that drive the global supply system. The series is intended to be technical and non-partisan. However, I promise that understanding how trade and outsourcing works won't weaken your next rant - it may give it a sharper edge.
This diary examines ways that purchasers deal with quality in sales transactions. Since most international trade is business to business (B2B) rather that business to consumer, the discussion focuses on that. I'm not expecting a big readership for this one, but I plan to link back to it from future installments. If you have questions, please ask them. I will either update, or add the answers to the later installments.
Future diaries in this series (titles may change):
- Limits of Inspection
- The New MakeOrBuy
- Management Control Systems
- Management System Standards and Auditing
- Quality Management Gone Wrong
- Management metastandards and globalization
- Do the Chinese have standards?
- Standards as a Progressive policy tool
... plus potentially many more
The Supply and Quality System
Taking delivery from a distant supplier is nothing new. Imagine what it must have been like for a Roman merchant to wait at the dock for his ship to arrive from Egypt. There were no UPS tracking numbers. He sent off an order by ship, then waited for weeks or months. When he spied the sail on the horizon, he knew the ship had survived, but what about the cargo? Did his supplier understand the order? Was the supplier able to fill the order? Did the supplier cheat him? Did the cargo survive the voyage? Those are the anxieties of every buyer since trade began. Fortunately, commercial buyers and sellers have evolved ways to increase their odds of success.
This diary is mainly a catalog of the various quality systems that are commonly encountered in trade and commerce. Think of it as an extended glossary. I will use these terms to examine more substantive operational, policy, and international relationship issues in later diaries.
If you break each transaction down to its basics, the following table shows the questions that need to be answered in order for any purchaser to feel comfortable that they will receive the goods they expect, together with the evidence that would typically provide reassurance.
These concerns apply to a system that, even in its most basic form, has a lot of moving parts. The following diagram is a simplified model of a typical supplier/purchaser transaction.
There are a lot of different ways to assemble and operate these pieces. The sections that follow look at some of the common variations on the typical buy/sell relationship.
Specifications and Standards
All transaction systems begin with a definition of the product. In business-to-business transactions (e.g., importing equipment), a written product specification will form part of the buy/sell contract. For a consumer purchase, the feature list on the side of the box or package may fill that role. The writing style for product specifications generally follows one of two distinct forms - or sometimes both together:
- Technical Specification - Technical specifications list the technical design details of the item being sold. Technical specifications describe the product in detail and may also include a 'method standard' that dictates how the product is to be made (e.g., how to make cheddar cheese).
- Performance Specification - The accelerated pace of technical change and the variety of potential designs has motivated buyers and sellers to adopt newer, more flexible specifications. Instead of describing the product and its method of manufacture, they set out the performance objectives (e.g., for a secure telephone) that the product must achieve - allowing the supplier to decide how best to achieve them.
Product specifications are generated by a wide variety of sources:
- Purchaser-written specifications - If the purchaser is knowledgeable (e.g., Proctor & Gamble, Hyatt, or Boeing), they may write a technical specification (e.g., to cater an ambassador's party) and demand that the supplier follow it exactly.
- Supplier-written specifications - When the supplier is more knowledgeable, the supplier may describe the product's features (e.g., a technical data sheet for an electronic device) and offer it to potential purchaser - take or leave it.
- Voluntary published standard - Anyone can publish a specification and call it 'standard'. If supplier/purchaser pairs somewhere decide that text of the standard is useful, they can elect to make it part of their contract. Since it helps if the publisher of the standard has independent authority or credibility, most voluntary standards are published by industry associations (e.g., Society of Automotive Engineers), non-profit organizations (e.g., American National Standards Institute), governmental agencies (e.g., AFNOR), and international standards publishing bodies (e.g., ISO).
- Mandatory standard - These are government regulations that apply with the force of law. By far the most common are standards that protect public or consumer safety (e.g., passenger restraints in automobiles). Any contract that involves the sale of these items must meet these standards first - before they worry about the specific needs of the two parties.
No one really knows how many published standards are in existence, let alone in use. Almost certainly, they are numbered in the millions (not counting versions in various languages). Privately written product specifications are uncountable. Regardless, it is hard to overstate their importance in facilitating the commerce that we take for granted. Without them, no light bulb would ever fit its socket. No PC computer would ever accept a peripheral. No food would ever be safe. They make our increasingly technical commerce possible.
Anyone who wants to influence the behavior of commerce (local, domestic or global) need look no further than the standards-making process to find a multitude of exceptionally powerful levers. This is a theme that will be evident throughout the diaries in this series.
Quality Management Systems and Strategies
Once a buyer and seller agree on the specification, their next task is to figure out how the seller can 'assure' the buyer that the delivered product will always correspond to their mutually-agreed specification. Over the centuries, a number of techniques have been developed.
Ownership: The simplest approach is for the purchaser to make it themselves, or closely related: to buy the supplier. This comes from the classic 'make or buy' decision that all purchasers must face. Do I make it myself? Do I buy my supplier? Do I merge? Do I sell my internal operation? Do I form a joint-venture? The issues are complex and endless. Lots of new twists are appearing. In fact, companies have become so creative that I am postponing detailed discussion of this topic to a later diary. For this diary, I'm sticking to the classic buy-sell transaction.
Suffice to say that the traditional distinction between suppliers and purchasers' roles has broken down. These shifts have a big effect on the distribution of capital and jobs, locally, nationally and internationally.
Supplier Reputation or Caveat Emptor: One of the most common arms-length approaches is for the seller to base their assurance on trust and a history of good performance in similar contracts. In effect, the supplier stakes their reputation that they will successfully deliver the product. This method is common with consumer goods. If you have ever bought a product from Sony, or an HP printer, or a pint of Ben & Jerry's Ice Cream, you probably did so because of the reputation and/or your previous experience with the brand.
Does the fact that you (or by reputation others) received good products in the past have any bearing on your chances of getting good product in the future? If you think about it, the answer is probably no - unless you know how the supplier was successful in the past. Maybe something has changed in their factory and their quality will soon fall sharply. If you don't know why they are good, how will you know when they are bad? Rationally, you should expect that defective products will eventually sneak through if you keep buying long enough.
Despite its limitations, this method remains popular because it's inexpensive. As a buyer or a seller, you don't have to do anything. Just keep buying and selling and hoping. As long as things go well, you're reassurance will grow - until something eventually goes wrong.
As consumers, we often rely on this mechanism way too much - especially when products come from global sources. We have found that the domestic companies that imported the wheat gluten that tainted the pet food knew next to nothing about the Chinese companies that supplied it.
Inspections: Inspections are one of the most misunderstood ways to ensure product quality. Typically, the supplier can inspect the goods prior to shipment (outgoing inspection) and/or the purchaser can inspect when they receive the goods (incoming inspection). From a technical and cost standpoint, it is usually better for the supplier to do the inspection. The supplier should have a deeper understanding of the potential defects, and if the supplier finds a problem, it can substitute or fix it and still ship on time.
When purchasers inspect, it is usually because they don't trust the supplier or they trust their own inspectors' competence more. Generally, purchaser-based incoming inspection is a 'necessary evil'. The purchaser must bear an inspection cost even if no problems ever show up and if a problem does surface, the purchaser must wait for a replacement to be shipped. The only thing the purchaser gains is confidence that defective goods won't be passed to its customers.
Inspection is widely used because it is obvious and easy. Most sales contracts put the onus on the supplier to inspect, with the purchaser reserving the right to make 'spot inspections' to keep the supplier honest.
The next diary in this series will look at the theory and logic behind product quality inspections. There are serious limits to what inspections can accomplish. Hopefully, it will resolve some misunderstandings about the nature of the 'statistical sampling' that is so often cited as being a part of quality and safety inspections.
Voluntary Quality Management System (QMS): Inspections can only find a defect after the product has been made, after all of the labor and materials have been invested. Products would be cheaper and better if the defects never occurred.
In and before WWII, manufacturing quality experts devised radical new approaches to achieve this utopian goal. The idea is to design and implement a production system that totally eliminates the sloppiness, inconsistency and uncertainty that permits defects to occur. The industry term for this is a 'quality management system' or QMS. In theory, if defects aren't possible, inspections will be unnecessary. A growing number of manufacturing companies are amazingly close to reaching this goal.
Many in manufacturing believe that a voluntary, supplier-driven QMS is the ideal solution to ensuring product quality (and safety, green operation, etc.). Unfortunately, for every company that pours its heart and soul into building a great QMS, there are probably 20 that make the claim, but not the effort. As a result, unsupported supplier claims for their QMS are highly discounted by purchasers.
The main exception involves a few companies that have taken a public lead in QMS innovation. Companies like Toyota (Toyota Production System) and GE (Six Sigma) are open about their approach and actively evangelize their features. The relentless advocacy tends to remove doubt that their commitment is real. Unfortunately, the list of these inherently credible QMS champions is very short. The vast majority of suppliers must earn their credibility from some other source.
Most of what we have learned about managing dangerous or difficult processes has come, not from academia, but from a small number of individuals, companies and organizations that were pioneers in their time. At one time or another, important ideas have come from the likes of Deming, Juran, Toyota, Panasonic, NASA, Xerox, Motorola and GE. Collectively, they have contributed greatly to our safety and quality of life. In this series, I will try to give them credit wherever I can.
Factory Certification to Purchaser-Specified QMS: As large companies recognized the value of the QMS, they began to demand them from their suppliers. Influential purchasers went further and publicly issued their own standard. These big purchasers expected suppliers to implement their QMS design and offer it for a periodic review.
This approach had one good feature and a lot of bad ones. The good feature was that the purchaser could impose a QMS design that would protect it from the defects it feared the most. The bad side included the cost, complexity and confusion that this approach created in the industry. Imagine if every purchaser wrote their own standard and imposed it (with factory visits) on each of their thousands of suppliers. The purchasers would spend a fortune on visits and the suppliers would be visited by multiple purchasers - each demanding that they follow a different QMS design. Ouch.
This model was prevalent in the 70s, 80s and early 90s - first in Japan and then in America. However, it is seldom seen today. Better approaches replaced it.
Public QMS Standard and 2nd Party (Customer) Auditing: The cost and complexity of the purchaser-written standard is reduced if the parties consent to adopt a published QMS 'metastandard'. The supplier's obligations are more generalized, but this approach eliminates the need to meet competing, purchaser-specific requirements. It's slightly weaker tea, but costs a whole lot less for everyone.
This model became popular with government purchasing groups (starting in the late 1950s) and industry associations in the 1980s. It started with products that had national or safety concerns (e.g., aircraft, drugs and military equipment) and spread to industries where everyone had roughly similar business needs (e.g. steel). Purchasers would write the published standard or regulation into every contract and purchaser-employed inspectors would visit the supplier's factory to verify that the required QMS was in use.
Publicly published management system standards were largely invented by the military procurement process. That approach, which dominated DOD purchasing for nearly 5 decades, proved to be another evolutionary blind alley. It was, however, a critical stepping stone towards the development of modern quality 'metastandards'.
QMS MetaStandard with 3rd Party Auditor and Certificate: In the early 1990s, a refinement of the QMS standard appeared. Using this standard, an independent auditor could replace the purchasers' site-visit teams. Instead of a gaggle of visitors, one team from a UL, KPMG, or a Lloyds Registry would conduct an annual or periodic audit. If you passed the audit, the 'registrar' would issue a certificate attesting to your compliance with the standard.
Anyone who received a certificate could demonstrate the reality of their QMS by faxing the certificate to potential purchasers anywhere in the world. The company could even put the registrar's logo in their ads - spreading the word even further. Customers could (and did) make possession of a certificate a qualification for bidding on contracts.
The idea of an independently-certified QMS also extended to companies around the world. More specifically, the ISO 9000 standard, first issued in 1987, has probably been implemented by a million firms worldwide by now (it was already up to nearly 800,000 in 2005).
I refer to ISO 9000 and its kin as a 'metastandard'. I will explain the theory behind these odd, new standards in a subsequent diary. For now, trust me that it is possible (within limits) to standardize important aspects of management behavior. I suspect this might excite some of the creative policy juices in a few readers.
I will refer often and in great detail to ISO 9000 in subsequent diaries, but not because I am advocating for or against its use. Rather, ISO 9000 is the most active and globally relevant of the management system metastandards. Its story contains a wealth of insights about the nature of management, management control and the prospects for achieving practical, widespread influence over management behavior.
Management System MetaStandards for other Issues: As buyers and sellers and national trade organizations have moved toward a global certification system for quality management, other organizational stakeholders have been working on their own assurance issues. In effect, they are defining 'quality' much more broadly. Is it a quality product if it was made with slave labor? Is it a quality product if a river had to die to make it?
Many of these initiatives are following trajectories that resemble, or directly borrow from, the quality management evolution. They incorporate aspects of published standards, independent auditors, and public certificates. A few of the more interesting initiatives are listed below:
The implication of this last table is one reason that I am writing this series. It is a key destination for the material I plan to cover. There is a real possibility that this new class of standards can influence the corporate and governmental behaviors that most concern many progressives. How strongly they impact those behaviors and how widely that effect will be felt remains to be seen. I suspect that it will depend at least in part on the level of technical understanding and insight that progressive activists can bring to the debate. I hope to further that possibility.
In this diary, I want to lay the groundwork for diaries that follow. I plan to link back to this diary, so it will can build as a reference. I would be interested in any comments, criticisms or questions that you may have. If you catch an error, I will update this diary. If you ask a good question, I will try to put a response in one of the forthcoming diaries.
Finally, a question to chew on ... and a poll.
Would you be more willing to buy from the [Chinese|Indians|Mexicans] if you believed that they adhered to the same management standards that we demand and expect from American firms?