A new Perspective on Cost versus Conformance Quality:

Achieving Conformance Quality as a Competitive Requirement

by Prevention Investments in the Long and by Rectification in the Short Run








Umit Akinc

Calloway School of Business and Accountancy

Wake Forest University

Winston-Salem, NC


(336) 758-5035










June 2001




This paper focuses, within the broader trade-off issue in competitive strategy, on the connection between conformance quality and the cost to achieve it.  The paper first suggests a model to analyze the strategic economic role of a firm’s quality policy.  It then articulates a new interpretation of the cost of quality models pioneered by J. Juran in the decade of the seventies. The new interpretation drives from two observations. These are 1) the distinction between the concepts of process conformance rate (capability), versus the market conformance rate that the customer experiences; and 2) different behavior of various costs, (particularly prevention costs), in the near term and the long term.  Although the interpretation based on these observations shed some light on the trade-off between cost and conformance quality, there is still no a priori argument to rule out the possible conflict.    



1. Introduction

The two decades or so starting with the late 70s may be characterized arguably as the era of renaissance of quality management.  During this time, product and service quality has occupied the minds of business practitioners and researchers probably more than any other topic.  Many organizations— both business and otherwise, have frequently attempted quality improvement programs and achieved various degrees of success. Concurrently, a voluminous literature— not all free from controversy, on topics ranging from mundane quality control charts to the role of quality in a company's competitive strategy has accumulated. Within a relatively short period of time the nature of quality management has undergone a transformation. Whereas only a few decades ago, product and service quality was something that belonged to the factory floor, hourly machine operators, and quality control technicians, recently it has been elevated to corporate boardrooms, strategic plans and company mission statements. According to Garvin [1987], until 1980s even the most noted quality experts such as Juran and Feigenbaum, wrote on quality management from a mostly technical perspective, rarely stressing its strategic impact and importance.

Literature on manufacturing strategy, since roughly 1970s starting with Skinner’s [1969] seminal work, recognizes quality, along with price, delivery performance and flexibility as capabilities on which a company can build its competitive strategy (for instance, see Stonebreaker and Leong [1994]).  Whether these capabilities are or are not subject to trade-offs has been, and is being hotly debated.  Two prevailing views are: 1) that there are inherent trade-offs among these capabilities as Skinner [1996] reiterates and clarifies his earlier work [1969] that a plant designed to excel in one of these capabilities may not be able to excel in the others. And 2) that it is plausible for an operation to improve on all the capabilities simultaneously as claimed by some recent studies, (Ferdows and DeMeyer’s [1990] ”sand-cone” theory, Hayes and Pisano[1996] and others).   Although this debate on the trade-off issue in manufacturing strategy is relevant to the main theme of this paper, divulging into an extensive review of manufacturing strategy literature is beyond its scope.  For an empirical treatment of the trade-off debate see, for instance, Safizadeh et al [2000]. 

This paper focuses, within the broader trade-off issue, on a narrower trade-off between cost and quality.  The existence of possible conflict between cost and quality has also been fervidly debated for a long time now. Unfortunately, however, the question of “is quality free?” has not yet been fully settled.  The dilemma stems from Juran’s [1951] well-known model of “cost of quality,” which seems to imply the existence of a minimum-cost quality level, which may occur at short of perfect quality.  This, in turn, has been an uncomfortable conclusion to accept for many authors, led by Crosby [1979], who regard quality that is short of perfect as completely unacceptable.  These authors have attempted, using various arguments, to show that indeed perfect quality and minimum cost levels are one and the same level. 

This paper attempts to articulate a new interpretation of the existing cost of quality models to bring a new perspective to the cost vs. quality dilemma.  The paper first explores the cost-quality connection, namely the economic role of a firm’s quality policy within the framework of its competitive strategy.  It then offers a new interpretation of the traditional cost of quality models based on two observations.  One observation deals with the distinction between two definitions of conformance: the conformance inherent to a process as opposed to the conformance that the customer sees.   The other is the behavior of various cost components in the short as opposed to long-run. This behavior is partly due to a distinction drawn between two types of prevention interference: day-to-day prevention activities whose results and costs are short-term and prevention investments whose results and costs span a longer period.  It argues that depending on which definition of conformance and what time frame is used, the traditional cost models may imply different things.  It shows that although the long-run behavior of the relevant costs tends to push the point of optimal quality towards “perfection,” there is still no conceptual a-priori argument to rule out possible cost vs. quality trade-off.  Even in the long-run, whether perfect quality will coincide with minimum cost point remains an empirical question to be answered on a situation by situation basis.

The rest of the paper is organized as follows. Section 2 presents a brief but critical review of the literature dealing with the issue of cost of quality; and undertakes a more in-depth discussion of Juran’s cost model and the literature championing the coincidence of minimum cost and perfect conformance quality.  Section 3 presents an original model of quality policy.  This model not only dissects the term quality into components relevant for the purpose of trade-off issues but also situates quality (with its long and short-run implications) in the broader context of a firm’s competitive strategy.  Section 4 presents a revised interpretation of the cost of quality model, based on two definitions of conformance and long and short-run cost behaviors.  Section 5 concludes the paper.




2. Relevant literature

Ever since Juran [1951] tried to cast quality as a management realm in terms of avoidable and unavoidable costs, the concept of cost of quality has received a great deal of attention.   Juran recognized three categories of relevant costs that vary with non-conformance rates.  Failure costs include internal failure costs such as rework, scrap and external failure costs such as recalls, warranty, litigation, and indirect (opportunity) costs such as loss of customer goodwill.  Secondly, prevention costs are incurred for activities and programs that aim at performing the required tasks correctly.  They include, among other activities, product and process redesign, employee training, preventive maintenance, process improvement by uncovering and eliminating root causes of non-conformance.  Thirdly, appraisal costs include the costs of sampling, inspection, screening/rectification (complete inspection and correction of non-conformances since the last ample either by replacement or rework), data collection, processing and analysis.

Juran termed the failure (non-conformance) costs as avoidable to lead managers to think about quality of conformance in more motivating cost terms rather than in abstract conformance indices.  He proposed his well-known model for minimizing total cost of quality as in Figure 1.   In this model the sum of appraisal and prevention costs, termed as unavoidable, is represented as a non-decreasing convex function of conformance rate.  This functional form implies that achieving higher and higher rates of conformance requires increasingly more stringent quality appraisal and/or more extensive prevention, causing these costs to climb at an increasing rate.  Likewise, the sum of internal and external failure costs, termed avoidable, is modeled as a non-increasing convex function of conformance rate implying that these costs climb at an increasing rate as conformance level worsens.

Figure 1 about here

If, indeed, the cost relationships are as modeled, and the overriding company objective is to maximize profits, then Juran's minimum cost conformance rate is on solid economic theory grounds.  This is the point where the marginal cost of conformance (prevention and appraisal) equals that of non-conformance (failure).  Unfortunately, however, this minimum cost point has a very unappealing feature.  It may, and most likely will, occur at a point where conformance is less than perfect.  This result is quite contrary to many of the maxims, principles, and advice from the majority of quality gurus who defend 'zero defects' as the essence of any sound quality policy.  In the current business atmosphere where quality is widely considered to be the core of any successful business/competitive strategy, a less-than-perfect conformance rate, albeit cost-wise optimum, is indeed quite difficult to accept.

A dilemma has then arisen between Juran's “traditional economic solution” and the appealing concept of  zero defects” as the basis of a firm’s quality strategy.  To reconcile this dilemma, many authors led by Crosby [1979] have attempted to show that the minimum cost solution should indeed occur at the perfect conformance point.  The prevailing argument in this regard is based on the possible shape and/or position of the prevention/appraisal costs.  Feigenbaum [1983] has argued for additional “indirect” failure costs, such as excessive inventories, poor design, loss of market share etc. which may result from poor quality.  Inclusion of these intangible costs tends to shift the failure costs curve towards the right (figure 1) pushing the minimum cost quality level closer to full conformance. 

Some other authors, in addition to the indirect costs, have argued for various and specific reasons for the relatively flat shape for the prevention costs.  See for instance, Diallo et. al. [1995], and Carr and Tyson [1992].   According to this view, if the prevention cost function is not heavily convex but sufficiently flat, minimum cost will be attained at full conformance level as in figure 2.  In the same spirit, Crosby himself claims that "prevention costs do not increase marginally, that no problems are diabolically difficult, and that systematic approaches to prevention of defects do not increase in cost."  This line of reasoning seems to resolve the dilemma in a rather neat way.

Figure 2 about here

The crux of this argument is based on an untested presumption however.  After all, the shape of the prevention cost, or all other costs for that matter, is an empirical question to be considered separately for each individual situation.  There is absolutely no a priori basis on which one can justifiably make such a blanket assumption.  In fact, if any argument can be made a priori, there is perhaps some theoretical defense for an increasing, convex prevention cost function as envisioned by Juran.  At any point in time, a process (e.g., a piece of equipment) may be associated with a set of problems (root causes) that prevents it from performing perfectly.  Prevention activities can be regarded as efforts to find and correct these causes.  If it is assumed that all the root causes are not equally difficult (or easy) to discover and correct, it, then, follows that the 'easier' root causes will be discovered and ameliorated first, making the next step of improvement harder to achieve.  This implies a convex cost behavior.  The plausibility of this assumption may be demonstrated by the Ishikawa or cause-and-effect diagrams (see for instance Omachonu and Ross [1994]), which are widely used to analyze process problems.  In these diagrams, some process problems are close to the backbone (primary), while other problems may be secondary, tertiary or even higher order.  The root causes at the end of longer causality chains are more difficult to discover and correct.

Narasiman, Gosh and Mendez [1994] model the impact of superior quality and price on the market share of a firm. This model offers indirect support for the hypothesis that cost and quality are not necessarily conflicting objectives.

Another cluster of studies has tried to reconcile the trade-off dilemma by the behavior of Juran’s original cost relationships over time.  Fine [1986] and in way of extension later, Marcellus and Dada [1991] have considered the effects of quality-based learning on the prevention costs using rigorous analytic models.  Their theory maintains that firms choosing to produce at higher conformance rates will experience faster learning thus faster cost reduction in their production activities.  This way, they see cost and quality as reinforcing rather than opposing objectives.  Prasad and Tyson [1995] synthesize the learning effects influencing the prevention costs, with the forces that competitors exert by improving their quality, on the failure costs of the company.  They show that though both of these forces will move the optimal quality level towards full conformance.  Interestingly, however, as quality approaches towards full conformance, the optimum cost could decrease as well as increase depending on the relative strength of these forces.  These works are closely related to one of the themes of the present paper which explores the same type of temporal phenomena at a more conceptual and a broader context.  This paper extends the scope of the analysis to distinguish between long and short-run implications of the dynamic behavior of all quality-related costs as a result of deliberate and economically justifiable management policy and not just passive result of learning phenomena.


3. Quality Policy and Competitive Strategy

Following the early lead of Skinner [1969] several important works have appeared in mid 80s (e.g., Hayes and Wheelwright [1984] and Buffa [1984]) which explored the role of manufacturing in a cogent competitive strategy.  The path these pioneers have opened has been traveled by a growing number of authors, some of which were mentioned in section 1. 

3.1 Structural, and Quasi-Structural Strategic Decisions. Hayes and Wheelwright [1984] characterize manufacturing as a “formidable competitive weapon” and define manufacturing strategy as a pattern of decisions over time by which the competitive factors of high quality, low price, delivery performance and flexibility are pursued.   Strategic decisions are generally associated with a long planning horizon, both in terms of carrying out such actions as well as realizing their outcomes.

Buffa [1984], on the other hand, categorized the major strategic decisions into a relatively small number of more specific realms: positioning the production system, capacity/location decisions, product and process technologies, vertical integration and suppliers, and strategic implications of operational decisions.  All of Buffa’s categories, except the last one, affect the very nature of the system by which a company creates value.  As such, these decision categories may be termed as structural-strategic decisions.   Strictly speaking, the last category, day-to-day operational decisions, when each occurrence taken separately, are not strategic: they do not have the properties that the structural strategic decisions have, such as a long-term horizon, pervasiveness, impact, etc.(Hayes and Wheelwright [1984]).   However, when viewed as a pattern of decisions over an extended period of time, they may have profound effects on the competitive and strategic outcomes of the company.  For this reason, we will refer to policies that guide repetitive operational (short run) decisions over an extended period of time as quasi-strategic.  For example, adding a new plant or expanding an existing one are clearly structural strategic decisions.  While a decision at a specific date regarding the inventory replenishment quantity of a certain component is not strategic, an inventory replenishment policy (pattern of operational decisions), based on a JIT philosophy has strategic importance and it is thus quasi-strategic.

3.2 Primacy of the Economic Objectives.  Although recent theories in strategic management consider a broader set of objectives than the mere profit motive which drives corporate strategy, (see for instance Kaplan [1990]), the priority of economic objectives is seldom challenged.  Therefore we start with the premise that most, if not all, companies are in business to make money in the short as well as in the long-run.  Most other stated or implicit goals can be traced to this objective.  A corollary to this premise is that the role of a quality policy, (or of any managerial endeavor for that matter,) in the competitive strategy of a company can only be assessed in term of its impact on the economic outcomes.  To be an effective and viable strategic weapon, the quality policy of a company cannot stand on its own but must be built on a solid economic base.  The policy must be more than a highly simplistic or evangelical recommendation: "achieve the best score on each of the quality dimensions," however defined.   A solid quality policy cannot solely rest on lofty moral or ethical maxims such as "do what is right;" or the righteous deed "to provide the best customer service;" or an impossible dogma "perfect quality in the short and in the long-run at all costs."   It must be based on a careful economic evaluation of a myriad of alternative actions and trade-offs in the long as well as in the short-run.

3.3 Quality Policy.  A quality policy that a company pursues as part of its competitive strategy may be defined as all decisions and choices that directly or indirectly affect a company's ability to offer to a target market quality product and services to enhance its economic objectives.  Thus a comprehensive quality policy has both long-run (structural-strategic) and short-run (quasi-strategic) components.  To be effective, both components, each affecting the economic objectives in unique ways, must clearly be aligned with the company strategy.  Both the long-run decisions such as the determination of product performance characteristics in the design process, and the short- run implementation issues, such as determination of operational parameters of a statistical quality control program, must be justified in terms of their economic consequences.

Figure 3 depicts various components of a quality policy and their relationship to Porter’s [1985] well-known “value chain.”  The components of a quality policy are on the right and can be broken into the areas of quality of design and quality of conformance  (Juran [1969]). 

Figure 3 about here

3.4 Quality of Design deals with the translation of perceived market needs into a set of design specifications and quality targets along various performance dimensions.  Product and service design is a critical strategic component whose impact on the company success transcends its quality implications.  It aims to deploy its competencies to best advantage by defining its markets and its competition.  The primacy of the economic objectives implies that the economically justifiable design must achieve a level on each performance dimension so as to equate the marginal worth of this dimension as perceived by the targeted market segment to its marginal cost.  Juran [1969] states this principle succinctly by saying: "We should put forth whatever effort is required to meet the needs of fitness for use.  We should avoid any effort which does not contribute to fitness for use."  There is general agreement for the existence of a trade-off between cost and quality of design. A product that achieves a higher performance on any dimension will cost more to produce.

            3.5 Quality of Conformance The long-standing debate on the possible trade off between quality and cost objectives pertains to the second part of quality policy-- quality of conformance.  This component includes all decisions affecting the frequency of adherence to the established design specifications and quality standards.  This is the definition of quality implicit in Juran’s [1951] cost of quality model.  Quality of conformance, as a management endeavor, in turn, can be broken into the areas of appraisal and prevention.  The short and long run cost behaviors of appraisal and prevention activities are quite different.  They affect economic objectives in distinct temporal ways.

Most appraisal activities deal with day-to-day execution of a quality control system, which most often consists of statistical process control, acceptance sampling and other related techniques to initiate a prevention intervention.  Appraisal may also include a level of rectification.  As will be seen in the next section, the rectification as part of the appraisal process has a paramount impact on the interpretation of the cost of quality models.

Quality control activities using these and similar techniques affect, on a daily basis, such costs as inspection, internal and external failure costs and such market objectives as conformance level as seen by the customer.  Strictly speaking, these decisions—frequency of inspection, decision rules as to when and what corrective action to take, etc. are short-term and non-strategic in nature, yet they are quasi-strategic in the sense that they have, over an extended period, strategic implications.

Prevention can be analyzed into two categories: On the one hand there are prevention activities which, may be considered as discrete investments whose outcomes are not necessarily confined only to the periods in which activities are undertaken but are realized over an extended period of time.  On the other hand, there are those day-to-day prevention activities whose costs and results are primarily confined to the current period and do not carry over to the future periods.

The model in figure 3 also makes a distinction between two concepts of conformance.  These are respectively, the process conformance, which is the level of conformance inherent to the process when it operates, as it normally should and the market conformance, which is the conformance that the customer sees after any appraisal/rectification activity.


4. Alternative Interpretations of the Cost of Quality Models

Both figure 1 (Juran solution) and figure 2 (alternative proposed by the flat prevention costs) are vague as to how the graphed quantities are measured.   Neither model specifies clearly where the conformance is measured— before or after appraisal/rectification takes place.  In both models ‘cost of conformance’ is plotted as a single curve representing the sum of prevention and appraisal costs; likewise, ‘cost of non-conformance’ is in aggregate terms and does not show how failure costs break in to internal and external components.  Furthermore, the time dimensions in which various cost components are measured are largely left unspecified.  The alternative view of cost of quality models is based on the distinction between the process and market conformance levels and between short and long run cost behavior.

4.1 Pre-and post-appraisal conformance--The fear to accept the less than full but minimum cost solution of figure 1, partly stems from the failure to distinguish between two definitions of conformance.  While the process conformance is inherent to the process and measures the level of conformance a process is capable of, prior to any appraisal/rectifying activity, the market conformance measures conformance after any appraisal/rectifying activity takes place.  In the context of the role of quality in competitive strategy, any reference to quality of conformance as a competitive priority must be made with market rather than process conformance in mind, as this is the level of conformance that the customer sees.  Clearly, any rectification of the non-conforming units, causes the process and market conformances to diverge unless they are both perfect.  While this distinction is quite important in product quality is it less so in service quality, since rectification in services is inherently more difficult. A poor service already delivered to a customer may not be easily rectified.  By aggregating appraisal with prevention, and internal with external failure costs together, traditional models blur this important distinction.  While prevention affects primarily the process conformance, appraisal/rectification my affect both process and market conformances.

If appraisal is intended exclusively for process control (detecting an out-of-control condition) and does not include any rectification, market and process conformance are identical thus the distinction is mute.  If, on the other hand, appraisal includes any rectification at all (as in many acceptance-sampling schemes), the market conformance (average outgoing quality) will generally exceed that of process conformance (incoming quality).  Neither model is clear as to which conformance is being optimized. With any rectification, as figure 3 illustrates, a portion of the process non-conformance is converted into another competitive priority, namely to costs associated with internal failure and appraisal/inspection/rectification activities.  Ideally, the extent of this conversion from one competitive priority (conformance) into another (cost) must be governed by its economic consequences.  Thus, when process and market conformance levels diverge due to any rectification activity, the relevant (conformance) quality policy becomes to determine not just the ‘optimal’ conformance level but also the ‘optimal’ appraisal/ rectification parameters—intensity and frequency of sampling, decision rules that trigger rectification etc. Unfortunately the effect of these two decisions on the costs are not independent.  For instance, increasing the aggressiveness of appraisal/rectification activity not only increases the ‘cost of conformance,’ but also affects the make up and possibly the sum of the internal and external failure costs at any given process conformance level. Thus a two dimensional graph is inadequate to represent the combined impact of two determinants of a quality policy—prevention on the one hand, and appraisal/ rectification on the other.

Though the two dimensional models such as figures 1 and 2 are insufficient to represent the prevention and appraisal/rectification components simultaneously, they may be interpreted as depicting the costs for a given fixed appraisal/rectification scheme. Accordingly, the optimal conformance level pictured is the process conformance optimized by varying the intensity of prevention activity, while the market conformance is somewhere between the optimal process conformance (corresponding to no rectification) and full conformance (corresponding to full census and rectification).  With this interpretation of figure 1, less-than-perfect process conformance does not necessarily imply shipping a certain percentage of poor products to the customers as generally implied by some proponents of model in figure 2.  If the external failure (including indirect) costs are sufficiently high, through a stringent inspection/rectifying process, a very high market conformance rate may be achieved regardless of the level of process conformance. Conceptually, for any given fixed level of appraisal/rectification activity there is a different total cost function with a possibly different optimal point.  The global optimal policy is that level of appraisal/rectification, which corresponds to the lowest of these optima. 

Figure 4. About here

Two such optima, which correspond to two rectification policies, are simultaneously pictured in figure 4.  As the rectification activity intensifies (drawn in bold curves), conformance costs (prevention plus fixed appraisal/rectification) increase for any process conformance level and shifts the curve to the left.  Also, assuming external failure to be more costly than internal failure, a more stringent rectification program catches a higher proportion of non-conformance, saving (higher) external, at the expense of (lower) internal failure costs.  At any level of process conformance, a more stringent rectification program thus results in a lower non-conformance cost, shifting the curve to the left.

Interestingly, as rectification activity intensifies the optimal process conformance level weakens. Although at a first glance this may appear to be surprising, a closer look reveals that this is indeed a plausible phenomenon-- market conformance somewhat protected by a more aggressive rectification, a lower process conformance becomes less expensive.  However, nothing can be said a priori about the relative magnitude of the optimal costs associated with, nor the inherent superiority of the two policies.  If the additional burden of intensifying appraisal/rectification is less than the savings in failure and prevention costs, the bold policy will have a lower optimal cost and vice versa.

An alternative way to make use of a two dimensional tool to analyze cost behavior is to plot the costs for a given process conformance level implied by fixed prevention activity as in figure 5.  When no appraisal/rectification takes place, market and process conformances coincide.  As appraisal/rectification intensifies, its cost climbs (drawn linearly here) and market conformance improves and finally reaches 100% when the intensity of appraisal corresponds to full census.  Meanwhile, as market conformance increases, the external failure costs decrease directly, whilst the internal failure costs increase, but more slowly. The optimal cost at this level of process conformance is the minimum of the U-shaped total cost pictured plus the exogenous fixed prevention costs. 

Figure 5 About here

When the process conformance (presumably achieved through more intensive prevention) improves to a higher level any desired market conformance can now be attained using a less stringent appraisal/prevention, reducing these costs and shifting the cost line to the left (not shown in the figure).  Obviously, better process conformance will lead to lower failure costs shifting these costs to the left as well.  If the total reduction in these costs more than offsets the additional prevention cost, the higher process conformance results in a quality policy with a lower overall cost.  Therefore the conceptual global is attained at that level of process conformance for which the minimum of the U-shaped cost curve (optimal appraisal/rectification) plus the exogenous prevention cost is the lowest.

4.2 Time Dimension of the Cost Components--Both models are static in that the time dimension of the costs is unspecified.  Do the functions represent hourly, daily, annual, or over some other interval average unit costs?  Whereas the failures, particularly internal ones, and appraisal activities result in immediate costs, at least some prevention programs (prevention investments in Figure 3) have a long-term nature. Thus their results and costs are not realized immediately. Design revisions, process alterations, supplier training and selection, employee-training programs do not result in cost and improved conformance rates coincidentally.  It is therefore somewhat misleading to compare them to activities resulting in immediate cost and benefits.  Such long term prevention interventions may or may not result in an immediate increase in conformance; and, like other capital investments, they may have impact beyond the period in which the activity is undertaken and, costs are incurred.   Representing the cost and benefit of such programs may not be a simple matter of comparison as implied by the models in figures 1 and 2. 

In addition to these long-run prevention programs, management may also have some short-term options, albeit limited, by which it can immediately affect the conformance rate.  Employing more refined and higher grade (and more expensive) raw materials, using the ablest and the most experienced operator, slowing down the process, performing routine process adjustments are examples of these short-run prevention measures.

To clearly differentiate between long-run prevention investments and short-run prevention activities, their results, and associated costs, we define the concept of the state of a process.   This is the level of conformance a process is inherently capable to achieve in the short-run (now).  Conceptually, this level may be associated with the set of remaining ‘root causes,’ which prevents the process from performing perfectly.  Prevention investments may be regarded as efforts to discover and eliminate some specific root causes once and for all.  Therefore, state of the process is a result of prevention investments up to the present without any extraordinary short-run prevention measures.   In the absence of any further preventive measures  (short or long-term), the average process conformance rate will correspond to the state of the process.  The actual conformance rate a process may achieve, in the short-term, will be around this base rate depending on the extent of short-run preventive interventions.   The short-run prevention cost can therefore be modeled as in figure 6, where the bend roughly represents the state of the process.  One can deviate around this rate by means of various short-run actions in either direction.

Figure 6 about here


As shown in figure 3, unlike short-run preventive measures, the long-run prevention investments alter the state of the process.  For instance, implementing an employee-training program, which may take several months, has the effect of correcting a root cause that keeps the process from performing perfectly. A long-run preventive program such as an employee-training program will benefit the process for periods to come without its cost being repeated each period by removing a root cause.  A preventive program of this sort improves the state of the process and moves the short-run prevention curve to the right since with an improved state of the process various short-run conformance rates can now be achieved more easily.  Therefore, the long-run prevention cost vs. process conformance rate can be depicted not as a single curve but by a family of curves as in figure 7. 

Figure 7 about here

The policy implication of this interpretation of the cost versus quality relationship is quite clear.  At any point in time (in the short-run), economically justifiable level of conformance is where the sum of short-run relevant costs is minimized as per Juran's classical model.  However, by investing in long-term preventive programs, the short-run minimum cost point will move to the right towards full conformance.  The primacy of economic objectives require that these investments not be made altruistically simply to improve quality but are to be justified on economic criteria such as return on investment-- just like any other investment.

The envelope of the short-run minimization points may now be termed the long- run prevention cost function.  The long run prevention costs in this model subsumes the concept of  'continuous improvement' as a dynamic process traced through implementation of economically justifiable pattern of decisions (investments) rather than a maxim that must be taken on faith.

Another important feature of this interpretation is that prevention programs’ cumulative effect over a long planning horizon dampens the climb of the long- run prevention costs.  Recognizing some prevention activities as investments whose outcomes occur over time lends more credibility to the relative “flatness” of the conformance costs as argued by some of the literature cited earlier.  The tendency of long-run prevention costs flattening over a period of time can be rationalized by reference to the cause-and-effect diagrams.  As economically justifiable prevention investments are undertaken and certain root causes are found and corrected over a time interval, other, more difficult-to-reach secondary, tertiary and even higher order causes move towards the middle line, making them more accessible (only through time) for discovery and abatement.  Also as the state of the process improves, the need for inspection/rectification declines as discussed in the previous paragraph; and of course, when and if the state of the process corresponds to perfect process conformance, the need to appraise/rectify will disappear altogether--the proverbial quality nirvana. 


5. Conclusions

The fundamental question is whether the quality related costs and quality, however defined, are compatible.  Although the conventional wisdom, mostly based on anecdotal evidence, is that they are, the answer does not seem to be an unqualified yes.  First of all, if by quality one is including quality of design, there is a clear trade off between quality and cost to produce.  A product aiming at higher scores on various quality dimensions (say, of Garvin's) will consume more resources and hence will cost more.  On the other hand, even if the quality is defined more narrowly as conformance to requirements (Crosby), quality and cost objectives reinforcing one another is a matter of time. Whereas in the immediate-run, given the state of the process, achieving conformance rates significantly beyond the state of the process may be very costly and thus the maxim "quality at all costs" may be economically indefensible as shown in figure 6.   Indeed, in the short-run, Juran's minimum cost conformance rate, which may fall short of full conformance, is rational.  The conformance that is being referred to is the process conformance; market conformance rate being protected through inspection/rectification depending on the relationship between external failure costs versus internal failure plus appraisal/rectification costs.  However, as prevention investments are undertaken on an economically justifiable pattern, the short-run rational level of conformance moves towards full conformance as shown in figure 7.  This march towards perfect quality along various short-run optima also lends a rational basis for the philosophy of continuous quality improvement.

Can one say that the dynamic view of prevention, appraisal and failure costs has explained away the dilemma of economic rationality and full process conformance?  Unfortunately, the answer is still no.  There is still no a priori basis to assume that the long- run optimum of figure 7 will occur at full process conformance.  To make this point, consider a process, such as in chip manufacturing, in which certain chemical processes are employed that are subject to irreducible uncertainties, possibly at the quantum physical level.  Due to the fundamental nature of the process, in the foreseeable future, no amount of prevention/improvement effort may be capable of moving the optimal point to full process conformance.  In a situation like this, even in the long-run the optimal yield of the process (process conformance) may be less than full.  However an economically justifiable appraisal/rectification program may be used to protect market conformance and thus competitive objectives.   If and when an entirely new and radically different process is discovered devoid of the inherent uncertainties, only then may the full process conformance be economically attained.  But in this case the attainment is not a result of any prevention/improvement activity but a fundamental redesign of the process outside the context of the prevention/appraisal and failure cost trade-off analyses.  The effect of such a radical discovery on the long-run conformance may be depicted as a sharp drop on the long-term conformance cost curve of figure 7.




Buffa, Elwood S.  Meeting the Competitive Challenge. Richard D. Irwin, 1984.

Carr, Lawrence P and Tyson, Thomas,  "Planning Quality Cost Expenditures," Management Accounting. October, 1992.


Crosby, Philip B.  Quality is Free. McGraw Hill, New York, 1979.

Diallo, A., Khan, Z. U., and Vail, C. F., "Cost of Quality in the New Manufacturing," Management Accounting, August 1995.


Feigenbaum, A. V., Total Quality Control. 3rd ed. McGraw Hill, N. Y. 1983


Ferdows, K. and DeMeyer, A. "Lasting Improvements in Manufacturing Performance" Journal of Operations Management. Vol. 9., No 2., 1990


Fine, Charles. H.,  Quality Improvement and Learning on Productive Systems. Management Science. Vol. 32, No 10, October 1986, pp. 1301-15.


Garvin, David A. "Competing on the Eight Dimensions of Quality," Harvard Business Review. Nov./Dec. 1987, pp. 101-109.  


Garvin, David A. Managing Quality. The Free Press, New York, 1988.

Garvin, David A. Operations Strategy: Text and Cases. Prentice Hall, Englewood Cliffs, N.J., 1992.


Hayes, Robert H., and Wheelwright, Steven C., Restoring Our Competitive Edge: Competing through Manufacturing. John Wiley & Sons, New York, 1984.


Hayes, Robert H., and Pisano Gary P. “Manufacturing Strategy: At the Intersection of two Paradigm Shifts” Production and Operation Management. Vol. 5, No 1, Spring, 1996. pp. 25-141


Juran, Joseph M. Quality Control Handbook.  McGraw Hill, New York, 1951.

Marcellus, R. L. and Dada, M. “Interactive Process Quality Improvement” Management Science. Vol. 37, No. 11, November 1991. pp. 1365- 1376.



Narasimhan, N., Gosh, S. and Mendez, D.  “ A dynamic Model of Product Quality and Pricing Decisions on Sales Response.” Decision Sciences. Vol. 24, no. 5 Sept 1993, pp. 893-908.


Omachonu, Vincent K, and Ross, Joel E., Principles of total Quality. St Lucie Press, Delray Beach, Fla., 1994.


Porter, Michael. Competitive Advantage. Free Press, New York, 1985.

Prasad, S. and Tyson, T.  “A Dynamic Model of Quality Cost Expenditures” Mid American Journal of Business, Vol. 10, No. 2 pp. 13-18.


Safizadeh, M.H., Ritzman, L.P., and Mallick, D., “Revisiting Alternative Theoretical Paradigms in Manufacturing Strategy,” Production and Operations Management. Summer, 2000, pp. 111-127.


Skinner, W. “ Manufacturing—Missing Link in Corporate Strategy,” Harvard Business Review, No 3, 1969 pp. 136-44


Skinner, W. “ Manufacturing Strategy on the “S” Curve” Production and Operation Management. Vol. 5, No 1, Spring. 1996. Pp. 3-14


Stonebreaker, P. W. and Leong, G. K.  Operations Strategy: Focusing Competitive Excellence. Allyn and Bacon, Needham Heights, Mass, 1994.