Dispute Prone Contract Clauses- A Basis for Operational Flexibility in Contract Administration

Iyer, K Chandrashekhar

Abstract

Contract terms and conditions play a dominant role in preventing and resolving disputes in construction contracts. However, in most construction contracts, a few clauses are regarded as dispute-prone. This paper presents a methodology for systematic identification of dispute-prone clauses in construction contracts using three multivariate statistical analysis techniques, viz. factor analysis, two group discriminant analysis, and multiple regression. The paper also tries to identify possible remedy for a reduction in the frequency of occurrence of disputes. A preliminary questionnaire survey of 72 clauses featuring in Indian Government construction contracts led to the identification 17 potential dispute prone-clauses. A more detailed questionnaire survey was then conducted for studying the 17 clauses. A total of 150 professionals (106 owner representatives and 44 contractors) responded to the questionnaire. The study revealed that dispute-prone clauses can be grouped in one of the four factor dimensions: owners' powers, contractors ' defaults, payment to contractors, and deviations and discrepancies. Clauses grouped in owners' powers were found to be the most dispute-prone from both owners and contractors points of view. Owner's decisions, governed by the power vested on him through these clauses and his fear of audit observations and objections, tend to be blind and rigid. If the owner's auditor had understood the operational difficulties of the project and been flexible in his attitude while raising the observations, the owner too would have the flexibility to mend his decision. This attitude can motivate contractors too to be reasonable in submitting the claim and this would, in turn, pave path of mutual confidence between them.

Keywords:construction contracts, disputes, contract clauses, statistical analysis, factor analysis, discriminant analysis, multiple regression, questionnaire, legal cases.

Introduction

Contract documents should play a dominant role in preventing and resolving disputes in the construction contracts by providing comprehensive coverage of all the items of work, duties and responsibilities of the contracting parties, and conflict resolution mechanisms related to all contingencies. However, in reality it is not so and a few contract clauses are regarded to be dispute-prone. Disputes may arise due to numerous reasons including unfair contract clauses allocating disproportionate risks on parties, ambiguity in contract clauses leading to varied interpretations and applications, discrepancies and inconsistencies among various documents, and conflicting interests of owners and contractors in the contracts.

Construction contract clauses predominantly represent owners' interests, to which the business hungry contractors agree, with a hope that enough ambiguity resides in the document to permit multiple interpretations (Ibbs and Ashley 1987). These clauses serve as proxy or pressure points while the parties assert their rights through claims. Situations become critical when the owner is a government organization and the owner representatives (the government officials) take decisions based on the content of the agreement rather than the intent, generally fearing audit objections, which are generally blind and rigid. A more flexible attitude of auditors that gives due considerations to situation and facts on the ground, may be a step in the direction of reduction of occurrence of disputes.

Literature Survey

Fox (1974) observed that owners prepared the contract documents with exculpatory clauses which dealt more with risks and uncertainties than with the specifications of the work. Thompson and Portis (1978), while discussing the history of evasive contract clauses from the pre-industrial revolution period, concluded that the use of exculpatory and evasive clauses by engineers to transfer all responsibilities to contractors was not a legally accepted method in contracting. Russel (1990) pointed out that competitive bidding systems were the main source of conflicts in contracts. Owners wanted the work to be completed at the lowest price with high quality and least risk of performance while contractors wanted to maximize their profits. Steen (1994) observed that even the most careful planning could not prevent disputes and suggested five steps to resolve construction disputes without going for litigation.

Thomas et al. (1990 and 1991) suggested methods for resolving disputes pertaining to notice requirements and oral changes. Thomas et al. (1995a and 1995b) have studied disputes arising out of defective specifications of construction contracts and on issues of compensation arising out of varied interpretation of completion of work respectively. These studies are related to the legal aspects of contract administration.

Ibbs and Ashley (1987) carried out a study on the impacts of construction contract clauses on four measures of project performance: cost, schedule, quality, and safety. They defined and examined 18 different clause families from 96 construction contract clauses. They found only three families among the 18 families to be significant in causing disputes. These families were 'change clause family', 'work scope definition clauses family', and 'project control clause family'. While this study gave an indication that the clauses were grouped together to represent families, no methodology for evolving the "clause families' was reported by them.

In India, Chandra (1984) has expressed the opinion that government contracts appeared to be totally one-sided and they are in favor of the government (the owner) allocating most of the risks to contractors. Due to this reason many of the clauses were not upheld by arbitrators or in courts of law. Moreover, a greater awareness on the part of contractors of their rights and remedies available through arbitration and also the growing tendency of playing safe on the part of owner representatives, have led to an increase in the number of disputes/arbitration cases (Agarwal 1991). Besides one-sidedness of the contracts, types of the contracts, documents forming the contract, interpretation of clauses, design changes and consequent effects, variation and extra items, risks and unforeseeable conditions, extension of time and escalation, had also been sources of disputes.

While these studies have provided useful information on particular clauses and the steps required to overcome disputes, no attempt has been made to study the nature of clauses in relation to the frequency of occurrence of disputes. This paper describes the use of three multivariate statistical analysis techniques for systematic identification of dispute prone clauses in construction contracts.

Research Objective and Scope

The primary objective of this study was to identify dispute-prone contract clauses from a construction contract. In general, construction contract clauses cannot be taken as completely independent of each other. Most clauses have a bearing on each other and they are correlated among themselves in some way or the other to form groups or families. So the first step to achieve the objective was to identify such groups of problematic clauses of a construction contract, which otherwise would seem quite unrelated to each other, and to identify their common underlying properties responsible for causing disputes. The next step was to identify the most dispute-prone clause(s) of a construction contract.

The Indian Government construction contracts used by Military Engineering Services (dealing with defence works) and Central Public Works Department (dealing with the public works), two quite old and large departments, were selected for study. The reasons for selecting these two departments are:

* These departments are almost a century old and have similar contract clauses with minor variations. The contract forms of these departments form the basis of many other contract forms followed by other the Central and the State Government Departments.

* These departments being large and pioneering departments of the Government of India, have experience in executing all types of construction projects, both in size and cost. Their establishments are spread all over the country.

Broadly, the methodology of the research study consisted of the following three steps.

* Data collection using a questionnaire survey approach,

* Analysis of data using appropriate multivariate statistical techniques, viz. factor analysis, two group discriminant analysis and multiple regression, and

* Legal case study.

Data Collection

Several professionals were approached for data collection and for gaining information from their past experiences, in the construction contracts. A few Government representatives declined to give interviews stating that they were bound by the Official secrets Act. However, they were ready to share their general views. Contractors, on the other hand, expressed their inability to respond stating that they retained records only for the current works and not for the completed works which were closed in all respects. Further, wherever data access was possible, it was found that organized information was not available for various contracts executed in the past.

Considering the above aspects, a mail questionnaire survey approach, where the respondents could be assured of the confidentiality of their responses, was used for data collection. This method also facilitated wider geographic coverage of the respondents. The questionnaire survey approach was adopted for data collection in two stages: (a) the preliminary stage, to identify the potentially problematic clauses, and (b) the main stage, to gather detailed information on identified problematic clauses and related issues.

Preliminary Questionnaire

The construction contracts considered for the study contain 72 clauses and many of them are not considered to be dispute prone. In order to identify only the potentially dispute-prone clauses for detailed study in the main questionnaire, a preliminary questionnaire was considered necessary. Since the questionnaire containing all the 72 clauses was quite lengthy and took more than two hours to complete, the survey was restricted to a smaller population. Of the 80 questionnaires sent, responses from 31 professionals (18 owner representatives and 13 contractors) were received and analyzed. Based on the rank order, indicating the decreasing order of contribution to disputes by various clauses, seventeen clauses falling in the top quartile were selected for further study.

Main Questionnaire

The main questionnaire was intended for a larger study and was quite exhaustive. It sought information on the respondent's professional experience and organization; issues related to construction problems (such as delays) disputes due to contract clauses, and several reasons associated with disputes, ill effects of disputes, and cost overruns. Since the scope of this paper is limited to identification of dispute-prone clauses in the context of their common properties and the impacts, an extract of the relevant portions of questionnaire is presented in Appendix I.

Questionnaire Responses

From the addresses available with the government offices, the Builders Association of India and through personal contacts, 380 professionals were identified and the questionnaires were sent to them in the first week of January 1995. all the professionals who were contacted had over 20 years of experience in the construction industry, and had direct association with defence/public works. The professionals included owner representatives (both active in service and retired) and contractors. A total of 150 responses were received which included responses from 95 owner representatives, 44 contractors and 11 retired government officials who were active as consultants with contractor organizations. The response rate was 39.5%.

The data from the responses were stored in a database and various analyses were performed using the SAS (1988) statistical software. Responses of consultants were not found to be significantly different, at 5% significance level using various statistical tests, either from that of owner representatives or of contractors. Also, since the sample size was small their responses were not considered for further analysis.

Identification of Underlying Properties of the Most Dispute-Prone Clauses

As discussed earlier, most clauses have a bearing on each other and they are correlated among themselves in one way or other to form groups or families. One method for identifying like-natured clauses to form sets or groups could be on the basis of personal experience. There would be chances of bias in such studies as identification would be based on the individual's personal experience. To overcome this problem, this study used collective experience of professionals having immense experience in contract administration through a questionnaire survey and identified sets of like-natured clauses from the point of their dispute proneness using powerful statistical techniques that eliminated bias almost completely. For identification of underlying properties of dispute-prone clauses, three multivariate statistical techniques were used: factor analysis, two-group discriminant analysis and multiple regression.

Factor analysis provides a method for combining variables from among several variables considered in a study, into sets of variables, called a 'factor' or 'factor dimension' (both represent the same nomenclature and hence the word 'factor' will only be used later in the paper). The variables present in a set represent some common properties among them, which may not be expressed or otherwise obvious in the beginning of the study. Factor analysis provides a method of combining the clauses into factors that are linear combinations of variables (the clauses) and independent of each other.

Two-group discriminant analysis was then performed to find out the relative importance of factors based on their discriminating powers between owner representatives and contractors. With the factors as defining characteristics and mean scores of the ill effects of disputes (Appendix II) as a criterion variable, multiple regression was then performed to find out the most problematic factor. The schematic diagram of the analysis along with the methodology for identification of dispute-prone factors is presented in Fig. 1. The various analyses are discussed in detail below.

Factor Analysis: Underlying Properties among the Clauses

Factor analysis is a powerful method of statistical analysis that aims at providing greater insight of relationship among numerous correlated, but seemingly unrelated, variables in terms of a relatively few underlying factor variate (Overall and Klett 1972 and Dillon and Goldstein 1984). This technique is extensively used by psychologists as a variable reduction technique. Many researchers from other areas including politics, sociology, economics, human-machine systems, accident research, taxonomy, biology, medicine, and geology have also applied this technique (Child 1990). In construction management, Maloney and McFillen (1995) have used this technique to identify various job characteristic differences between union and nonunion workers.

Factor analysis is usually employed for the following four purposes (Hair et al. 1990):

1. To identify a set of dimensions that are latent (not easily observable) in a large set of variables.

2. To devise a method of combining or condensing large number of people into distinctly different groups within a larger population.

3. To identify appropriate variables for subsequent regression, correlation or discriminant analysis from a much larger set of variables.

4. To create an entirely new set of a smaller number of variables to partially or completely replace the original set of variables for inclusion in the subsequent regression, correlation or discriminant analysis.

In the present study, factor analysis was employed for the applications 1 and 4 above. Important technical terms discussed to interpret the results of analysis are explained in Appendix III for greater clarity to readers. The terms, 'variables' and Observations' referred below are the 'seventeen clauses' and the 'responses of the individual respondents' respectively.

The total sample, i.e. responses from both owner representatives and contractors to all seventeen clauses for the question, 'frequency of occurrence of disputes', was subjected to factor analysis and four factors were extracted using the minimum proportion criterion. Initially, the extracted factors, which were all orthogonal to each other in nature, were not amenable to interpretation. Therefore an oblique rotation of the reference axes, called varimax rotation, was performed. The derived factors and their corresponding loadings were then obtained. A summary of the rotated factor loadings matrix is presented in Table 1 under the column 'All respondents' responses.' Factor loadings > or = 0.4 are only shown in this Table. The four factors explained 57.3% of the total variance. Further, the communalities of all the variables were found to be much greater than 0.3, hence the factor model was reliable. Meanings assigned to these factors are given below.

Factor-1

The first factor, Factor-1 has high loadings for five clauses viz., Owner's lien on withheld amount', 'post audit and technical examination', 'recovery of owner's claim from any sum due to contractors', 'final and binding power' and 'where no specification exists'. The clause, Owner's lien on withheld amount' describes that whenever any claim arises out of the contract against the contractor, the owner has the authority to withhold a sum of money from any sum payable to contractor under the same contract or from the security deposit of the contract. Further, this clause also states that the owner will enjoy the lien on the withheld amount till the claim is settled through arbitration or in court of law, and the contractor shall not have any claim for interest or damages whatsoever on any account in respect of such withholding or retention under the lien.

The 'post audit and technical examination' clause explains that the owner (the Government) has the right to conduct audit and technical examination of the works and the final bills of the contractor even after payment of the final bill. In case any discrepancies are noticed during technical examination of the work, a recovery towards noncompliance of specification is recommended by the examiner and the contractor shall pay the same.

The 'recovery of owner's claim from any sum due to contractors' clause describes that in case sufficient amount of the owner's claim, including suggested recovery on account of defective work or overpayment, is not available in the contract, the claim can be recovered from the running bills of the same contract or bills of some other contracts (which may be with other organisation or employer), or from any other source. The source includes the banks or the financiers to the contractor.

The 'final and binding power' clause discusses the owner's power to give decisions, which is an administrative authority of owner, which will be neither open to arbitration nor any court of law. Yet it is found that these decisions are being challenged both through arbitration and courts of law. This clause is applicable for : levying compensation on contractor in the event of delay in completion of the work, deciding the quantum of recovery while accepting sub-standard work (but sound structurally), and in many other situations. Various situations which lead to disputes where this clause is the central point, will be discussed in the section on legal cases later in the paper.

The the where no specification exists clause describes that if there does not exist specification in the contract document for a particular work, the Engineer-in-Charge shall have the authority to decide the specification and get the work executed.

The common property through which the above five clauses are linked to each other is the Owner's power'. However, if factor analysis had not been performed, the five clauses appear to be addressing different aspects of a contract on individual basis. For example, while Owner's lien...' describes withholding of claim amounts, 'recovery from...' discusses the mode of realisation of the claim amounts from other sources, if sufficient money is not available in the current contract. As for underlying property causing disputes under these clauses, they appear to relate more with payment to contractor. 'Post audit and...' addresses the quality check of constructed facility or auditing of bills to check any overpayment. 'Final and binding...' and 'where no specification...' describe necessary administrative measures required to be taken by an owner representative to achieve the objectives of the contract. However, the five clauses together explain the powers of the owner representatives of factor-1 which they could exercise on the contractors in the event of necessity. Hence, the factor was named as Owners' powers'.

Factor-2

In Factor-2, the clauses appearing are: 'suspension of work on default of contractor', 'time, delay and extension, and compensation for delay', 'completing unfinished work of contractor under risk and cost of original contractor due to defaults of contractor', 'rescission of contract on default of contractor' and 'foreclosure of work when it is no more required by owner'. all these clauses except the last one, explain various actions to be taken by the owner representative in case of contractors' defaults. Hence, the factor was named as 'Contractors' defaults'.

Looking at the descriptions of various clauses of the above two factors, all clauses appear to discuss owner's decisions or owner's powers in one way or the other. But there exists a distinct difference between the sets of clauses in the two factors which can be observed through a close scrutiny of the clauses. While the clauses of Factor-2 state various types of actions to be taken by the owners on the contractors' defaults, clauses of Factor-1 discuss mostly the owner's powers as a corrective measure of their own faults. For example, Factor-2 clauses state that: if the contractor delays..., or if the contractor fails to complete or rectify..., etc. Whereas from Factor-1 clauses it could be noticed that: if any overpayment is noticed at a later date..., or owner has the right to audit and examine the work after final bill is paid (the work would have actually been approved and accepted by one of the owner representatives) and if any recovery is noticed..., etc. The distinction between the two types of owner's authorities discussed above is not obvious and can easily lead one to believe that all clauses possess the same property of owner's decisions, if no rigorous analysis had been carried out.

Factor-3

The Factor-3 contains clauses: 'Foreclosure of work when it is no more required by owner, with due compensation for the work executed/material brought to site for execution of work, but without compensating for the loss of profit that the contractor might have earned from execution of full work', 'pricing of deviation/extra items.,', 'compensation for price escalation using price indices' and 'preparation and payment of final bill' to contractor. all these clauses explain methods of pricing, interval of preparation of bills and various checks for payment to the contractor. Hence the name 'Payment to contractors' was assigned to this factor. Incidentally, the clause, 'Foreclosure of...' appears to be associated more with Factor-1, the owner's power. This clause states that if at any time after commencement of the work, Government shall, for any reason whatsoever, not require the whole work as specified in the contract, the owner representative will give notice in writing and determine (i.e., foreclose) the work. This clause further states that the contractor shall not have any claim to payment of compensation on account of any profit or advantage that the contractor may have derived from the execution of entire work. However, the contractor would be paid for the items of work carried out by him at the tendered rates and for all materials brought to site for execution of work. Hence the appearance of this clause in Factor-3 gives an insight that the consequential effect of foreclosure, the loss of profit/payment to contractor is the main cause for disputes rather than the foreclosure by the owner. Similarly, 'pricing of deviation...' clause appears to be more associated with Factor-4 where other aspects of deviations have predominantly appeared. However, its appearance in Factor-3 indicates that the contentious issue behind this clause is 'payment to contractor' which is suspect to be subjective by both parties and finally leads to disputes.

Factor-4

The clauses dominating in the last factor, Factor-4 are 'subsoil data: owner does not take responsibility of correctness of data and contractor shall himself explore thoroughly at his own cost, before quoting', 'interpretation of contract in cases of ambiguity in various documents forming contract', 'deviation in foundation items' and 'deviation in superstructure'. all these clauses relate to deviations and discrepancies in a contract. Hence, the name 'Deviations and discrepancies' was assigned to this factor.

From the above discussions, it is clear that if the clauses are to be grouped based on some common properties, and without any rigorous analysis like factor analysis, there may be many different combinations. Each property can subsequently be claimed to be responsible for disputes. Moreover, because of the interrelationship among various clauses, there are also possibilities of simultaneous appearance of the same clauses, in several groups, e.g., pricing of deviation can be cited in Owner's power', or 'deviations and discrepancies', or 'payment to contractor'. Also it would be difficult to gauge the degree of predominance of individual clauses in defining the particular property. Whereas the factor analysis helped in evaluating the significance of individual clauses through their factor loadings in defining particular property-the factor. This analysis also helped in forming distinct groups with only one clause overlapping two factors.

Factor analysis was again performed on the seventeen variables for owner responses and contractor responses separately. Four factors were extracted in each case and then subjected to varimax rotation. Factor loading (loading > or = 0.4) of various variables (i.e. clauses) obtained after varimax rotation, for both owner and contractor responses, are summarized in Table 1 under the columns Owner Responses' and 'Contractor Responses' respectively.

The dominant variables in Factor-1 in owner responses were found to appear in Factor-3 in contractor responses and the dominant variables in Factor-1 in contractor responses appear in Factor-2 in owner responses. This meant that the clauses, which were dominating in owner responses (Factor-1) and perceived to be contributing the most to common factors (i.e. explaining maximum variance among all extracted factors), were less dominating (Factor-3) in contractor responses. Similarly, the set of clauses (Factor-1) dominating in contractor responses were not so dominating (Factor-2) in owner responses. Although this implied that differences in perceptions existed between owner representatives and contractors about the extent of disputes caused by individual factors, this could not be conclusively stated without proper statistical tests.

Further, the factor analysis did not indicate which factor influenced the most in 'the occurrence of disputes'. It indicated only the grouping of like variables in a given factor, based on their concomitant variation among the variables. It was possible, therefore, that the first factor, in spite of explaining the maximum variance among the four factors, was the least dispute-prone and some other factor was the most dispute-prone.

Incidentally, each of the factors identified above has been discovered to be dispute prone by earlier researchers based on their professional experiences. For example, Chokshi (1978), Agarwal (1991) and Roy (1991) had pointed out that no finality in owners' decisions, or improper utilization of owners' powers were great sources of disputes. While discussing various reasons for project overruns and disputes, Jergeas and Hartman (1994) and Saluja (1984) identified contractor's defaults as the primary cause for most overruns and disputes. Dennis (1991) maintained that if contractors' eventual payments were secure, then he was not prone to make the worst of any problem. Similarly, many other researchers have identified deviations and discrepancies as the root cause of most disputes (Fox 1974, Ibbs and Ashley 1987, Rau 1993).

Hence, in order to assess the dispute-proneness of individual factors, further tests were necessitated. Factor scores were then calculated from the factor loadings matrix of rotated factors and employed in performing discriminant analysis and multiple regression as discussed below.

Discriminant Analysis: Differing Perceptions

Discriminant analysis is a statistical technique for classifying individuals or objects into mutually exclusive groups on the basis of a set of independent variables. Several applications of this technique have been reported by researchers in the areas of psychology, medicine, marketing research and business (Dillon and Goldstein 1984; Overall and Klett 1972; Gau 1978). Hair, et al. (1990) have pointed out the following applications of discriminant analysis.

1. To determine if statistically significant differences exist between the average score profiles of the two (or more) a priori defined groups.

2. To establish a procedure for classifying statistical units (individuals or objects) into groups on the basis of their scores on several variables.

3. To determine which of the independent variables account for the differences in the average score profiles of two or more groups. In other words, this analysis reveals which of the specific variables in the profile account for the largest proportion of inter-group differences.

In the present study, two-group discriminant analysis was performed for the applications 1 and 3 above. Generally, the differences in opinions between two people about any matter can be regarded as sources of disputes. Accordingly, the objective of applying this technique in this study was to determine if there existed significant differences of opinion between owner representatives and contractors about the factors and if differences existed, which of the factors accounted for the differences?

The computed Z value for the model was 6.83. The Z follows an F distribution with p and n^sub 1^+n^sub 2^-p-1 degrees of freedom and its computed value was found to be greater than the theoretical F value (2.44), for the given degrees of freedom at 5% significance level. Hence, the model was statistically significant which indicated that the factors (the explanatory variables) used in the model adequately explained the group differences.

The results of the analysis (Table 2) indicated that Factor-2 discriminated the most, followed by Factor-3, in the responses of owners and contractors. These two factors revealed about 75% discrimination exhibiting maximum differences of perceptions about these two factors, between owners and contractors. If differing perceptions were the criteria to select the most dispute causing property of the clauses, the obvious choice would be Factor-2 followed by Factor-3. On the other hand, the Factor-1, which dominated among all factors explaining maximum variance, was found to possess the least discriminating power explaining only 4.2% discrimination. This percentage exhibited least differences in perceptions between owner representatives and contractors (or in other words, the similarity of opinions) about Factor-1.

At the end of the factor analysis the question remaining was : whether Factor-1 explaining the maximum variance was the most dispute-prone or the least dispute-prone? The discriminant analysis indicated that the Factor-1, explaining the maximum variance was the least dispute-prone. Following the similar logic, Factor-4, which explains the minimum variance among the four factors, should have been the most dispute-prone followed by Factor-3 and Factor-2. However, the results of the discriminant analysis were different. Hence before concluding about the dispute proneness of factors, further probing was required. Multiple regression was used for this purpose.

Multiple Regression Analysis: Disputes Versus their Ill-Effects

Regression analysis is primarily concerned with estimating and/or predicting the population mean values of a dependent variable on the basis of the known values of one or more explanatory variables. This technique has been used extensively by researchers in the area of sciences, social sciences, economics, management and engineering. In this study, multiple regression analysis was used for the limited purpose of identification of the factors most responsible for occurrence of disputes in construction contracts.

Factors were considered as the explanatory variables and the effects of disputes (Appendix II) as the criterion variable in the multiple regression. Factor scores, obtained after varimax rotation, represented the values of factors, and the mean scores of all eight effects of individual responses were considered as the corresponding values of the criterion variable. The results are summarized in Table 3.

It was found from the parameter estimates (Table 3) of the regression equation that Factor-1 was the most significant among all the factors with coefficient of variable as 0.13142 and a p-value (PoItI) = 00 .0 05480 .0 But since the F value was low (Pr>F = 0.1619), the regression as a whole could not be considered significant.

On further investigation, it was found that the criterion variable, which represented the ill effects of disputes through mean scores of responses to all the eight effects listed in the questionnaire, did include a few effects which were not considered to be caused by the disputes. The individual mean values of effects numbered 3, 4, 5 and 6 (Appendix II) were either significantly less than 30 .0 0 or significantly not different from 30 .0 00 .0 This indicated that respondents were not in agreement with the facts that these effects were caused by the disputes in construction contracts. Hence, discarding responses to effects numbered 3, 4, 5 and 6, another iteration of multiple regression was performed.

Several iterations of stepwise regression were carried out considering fewer factors at a time. The important results are summarized in Table 4 as cases I to IV. Elimination of scores of four effects (effects numbered 3, 4, 5 and 6) from the criterion variable scores did improve the regression with regard to both the R^sup 2^ value and the F value. However, the p-value (Pr>F) was observed to be greater than 0.05 in all the cases studied in the stepwise regression (Table 4). Hence, none of the regressions could be considered significant at a 5% significance level whereas a few were significant at a 10% level.

Further, R^sup 2^ values in all cases, including the case described in Table 3, were found to be quite low. Normally three reasons are attributed to low values of R^sup 2^: noise factor, sampling error and variable selection. Explanation of these reasons with reference to the present study could be as follows.

Noise factor: Since several iterations of stepwise regressions were performed, keeping a few variables at a time and the R^sup 2^ value had been consistently low, the influence of noise factor was ruled out.

Sampling error: Possibility of this error could not be ruled out as responses from professionals having varied background, in terms of types and sizes of projects they had executed, were sought. However, since the sample size was large, the effect of this error may not be significant.

Variable selection: probably this reason was the most responsible factor for low R2 value. This may be due to certain key variables missing in the study. From the low R^sup 2^ value it was concluded that the criterion variable, represented by 'ill effects', depended not only on the disputes due to contract clauses but also on other reasons (characteristic variables), including disputes, due to other causes than only the clauses, which were not included in the study. This could be considered as a limitation of this study.

Since this technique was used for limited application of identification of most significant dispute-prone factor from among the four factors and not for carrying out any advanced predictions, the results of analysis at 10% significance level, were-relied upon. Accordingly, based on the t-values and coefficients of variables of regression parameters, it was concluded that Factor-1 had the greatest potential to cause disputes among all factors.

It can be further seen from the summary results of multiple regression given in Table 4 ( case I) that Factor-2 and Factor-3, which explained the maximum differences between owner and contractor responses, are not so significant (Col 5) in causing disputes. It is Factor-1, which exhibited the least differences between owner and contractor responses, that is the significant factor in the regression (Col 5). The two analysis techniques, the two group discriminant analysis and the multiple regression, together lead to conclude that both owner representatives and contractors have collectively expressed that Owner's power' is the most significant in causing contract dispute. Also based on coefficient of variables in case I regression (Col 3 in Table 4), it can be concluded that the next important factor in the regression is Factor-2, followed by Factor-3 and Factor-4.

Legal case Study

Most construction contracts in India have arbitration agreements for settlement of disputes. Since the arbitration awards are not publicized due to ethical reasons, they are not available for study. However, these awards can be challenged by the aggrieved parties on specific grounds in the courts of law (the parties here are not necessarily the contractors alone, but they can also be the owners). Again, the parties, if dissatisfied with the judgment of the court can file appeal in the next higher court with special leave from the lower court, up to the Supreme Court (the Apex court of the land). The law reporters publish only those cases which are heard and settled by the Supreme Court, or High Courts. Hence, the number of cases pertaining to construction contract disputes appearing in the law reporters is very few when compared to actual number of disputes arisen.

Moreover, these courts being appellate courts deal only with questions of 'law' (i.e. if any error of law has crept in the previous judgment for which it has been appealed) and they do not deal with questions of 'fact' or 'evidence'. Hence the cases are not easily comprehendible to engineers and many times the judgments appear confusing and very often it is difficult to trace the actual clause(s) dealt/ discussed in the specific cases. However, with the background of the survey analysis results, which helped in a great way to understand the court cases more clearly, 42 cases were studied. Though the scope of the questionnaire survey was restricted to only two departments, CPWD and MES, case studies were not confined to disputes arising in the CPWD or MES contracts only. They included other Central and State Government organizations. Most of the cases studied were reported in the Arbitration Law Reporter (Arb L R) between 1983-93.

It was found from the case studies that disputes connected with decisions under 'final and binding power' were the most frequent (18 times) followed by 'time, delays and extension' (13 times), 'rescission of contract' (11 times: includes completing unfinished work of contractor under "risk and cost' of contractor (5 times)), finalizing rates of 'deviation and extra items' by owner representatives (8 times) and definition of deviation limit/scope of work (4 times). It should be noted here that more than one clause were invoked simultaneously in some specific court cases and therefore, total count of invocation of clauses would not match with the number of cases studied. For example, when there were delays in completing work caused by the contractors, the owner representatives invoked one or more of the three clauses: 'levy compensation' for delay, 'rescission of contract', and 'risk and cost' for getting the remaining work completed through other contractors under the risk and cost of the original contractor.

The clauses 'time, delay and extension, and compensation', 'rescission of contract', and 'completing work under risk and cost', although actually dominate in Factor-2, as given in Table 2, the 'final powers' prerogative exercised by the owner representatives has governed decisions which have led to disputes regarding the three clauses. Similarly, on a few occasions, owner representatives using their final powers, reduced rates of certain items that did not conform to contract specifications. Then contractors raised disputes claiming that since mutually accepted substituted items were used, the recoveries were irregular and illegal. Although such disputes are apparently related to Factor-3, 'payment to contractors', or Factor-4, 'deviations and discrepancies', their root cause has been traced to be the decisions taken by the owner representatives under their 'final powers'.

Interestingly most disputes, directly or indirectly, revolved around decisions taken by owner representatives under their final powers, which were neither acceptable to contractors nor referable to arbitrators or any court of law. It was confirmed from the court judgments where several appeals pertaining to disputes arising out of decisions of owner representatives under their final and binding power were rejected. This further indicates that although the dispute-prone clauses have appeared predominantly in Factor-2, Factor-3 or Factor-4 (based on the selection criterion of factor loading > or = 0.4), they do possess some of the properties exhibited by Factor-1, i.e. Owners' powers', and contribute to factor score of Factor-1. Hence, the frequency of invoking certain clauses comprising a factor cannot be construed as being indicative of their being most dispute-prone. This fact is further strengthened from the results of the regression analysis, which showed that Factor-2 or Factor-3 or Factor-4 were not significant either in their coefficient values or t-values. However in case-I (Table 4), they collectively improved the characteristics of Factor-1, both in its coefficient value and t-value. Hence, it could be concluded that the latent property, Owners' powers' of the problematic clauses represented by Factor-1 is the main source of most disputes and the 'final and binding power clause' is the most dispute-prone clause.

The legal case study ruled that 'final and binding power' clause is not defective in its wording, but it is the subjective decisions of owner representatives due to 'their lack of knowledge regarding the legal implications of their decisions' or 'playing safe attitude to avoid criticism', which is the root cause of disputes in construction contracts.

Operational Rexibility in Contract Administration

Subsequent to the legal case study, a few professionals having a large experience in construction industry were approached to know their reactions on the findings. They also accepted the conclusions to be right and they added that the owner's or the contractor's attitudes to distort the language of the contracts and interpret for their personal gain (commonly called as reading between the lines) and their adamant and rigid attitude in going by the 'content' of the contract rather than 'intent' of the contract was the main source of dispute. These disputes remain unresolved for decades. If a good contract administration guidelines were prepared and implemented and the parties tried to look beyond their personal gains and understood the 'intent' of the contract, then there would be drastic reduction in the occurrence of disputes. Contact clauses should be regarded only as the tools at the hands of operators. It is the parties who are the operators and handle them. Having recognized that certain clauses are dispute-prone, these clauses should be invoked and used sparingly. One should adopt other possible methods of achieving project objectives and exercising of powers, vested in owner through these clauses, only as a last resort. This can be achieved only if flexibility in administration of contract is adopted. If this flexibility is adopted, it would also build confidence in the minds of contractors and motivate contractors to be reasonable while submitting their claims. Thus the operational flexibility brings the flexible attitude of parties and paves a path of win-win situation between the two parties of conflicting interests. Comments of the professionals matched with the findings of the authors in a paper (Iyer et al. 2001) where they have addressed problems arising out of unreasonable or rigid attitude of parties.

Summary and Conclusions

The present study aims at developing a systematic and quantitative procedure for identifying dispute-prone clauses from a construction contract through recognizing the latent underlying properties among them. Under the Indian conditions of contract administration, it was found that organized data of completed projects were not available for the study, either with owner organizations or with contractors. A questionnaire survey approach was therefore, considered for data collection. Based on a preliminary questionnaire survey, 17 potentially problematic clauses were selected from among 72 clauses. A methodology, using three multivariate statistical analysis techniques, viz. factor analysis, two group discriminant analysis and multiple regression was developed to analyze responses pertaining to the 17 clauses in relation to their impacts on occurrence of disputes.

Through application of factor analysis 17 clauses were transformed into four independent factors: Owner's powers', 'contractors' defaults', 'Payment to contractors' and 'Deviations and discrepancies.' These factors indicate the underlying properties of dispute prone clauses. To establish relative importance of factors as determinants of the occurrences of disputes, two-group discriminant analysis and multiple regression were performed using factor scores obtained from factor analysis. While discriminant analysis indicated that the two factors, 'contractors' defaults' and 'payment to contractor' exhibit maximum differences between owner and contractor responses, the multiple regression established that the factor, Owner's power' was the most significant in causing disputes. Interestingly, both owners and contractors have expressed similar opinion about the factor, Owner's power' which can be seen from the discriminant analysis results. Appearance of Owner's power' as the most dispute-prone factor in lieu of Factor-2 or Factor3 also demolished the general myth that differences in perceptions is the root of all disputes, at least in the context of the Indian contracts.

Moreover, the above results being the outcome of collective responses from 150 professionals having long and rich experiences, and application of multivariate statistical techniques significantly eliminated personal bias in the results. Following the survey and analysis, legal case study of 42 cases pertaining to construction contract disputes arising out of contract clauses and personal interviews with seasoned professional were undertaken. This gave greater insight to the findings of above stated methodology and established the reliability of analysis results. It was found that most of the disputes were centered around the clause, 'final and binding powers'. The decisions given under this clause are final, binding and conclusive on both parties and neither referable to arbitration nor to any court of law and, accordingly, appeals pertaining to disputes arising out of decisions under this power are rejected by courts.

Further probing into the cases revealed that the subjective decision of owner representatives due to their 'lack of legal knowledge regarding the implication of their decisions' or 'playing safe attitude to avoid criticism', and the rigid attitude of parties to contract are the root cause of disputes. The clause, 'final and binding powers' is not defective in its wording. Hence good contract administrative guidelines are required to be formed to bring in more operational flexibility in the contract administration and implementation. Owner representatives should be trained to take proper administrative measures to prevent/resolve dispute and both parties should be encouraged to be more flexible in their attitude and stand.

Finally, the results of this study also demonstrate that even if documented information of completed projects are not available for study, the methodology employed in this research could be incorporated in identifying dispute-prone clauses from any contract worldwide.

Acknowledgement

The authors thank all respondents for sparing their valuable time for reeponding to the questionnaire. The authors also thank Prof .M. Ranganathan of the Department of Commerce, Madras University, Madras (India) for his valuable suggestions during the course of this study.

References

Arbitration Law Reporter, Arbitration Law Reporter, Delhi (India).

Agarwal K.N. (1991). Challenging an arbitration award before the court. Seminar on Contract Management & Arbitration in Civil Engineering Projects, The Institution of Engineers (I), New Delhi, 22-23 September, IV-15 - IV-21.

Central Public Works Department. Item Rate Tender & Contract for Works: CPWD - 8. Ministry of Urban Development, Government of India.

Chandra H. (1984). Government contracts and tendering procedures. all India Seminar on Construction Contracts - Contribution by Contractors, Architects, Consultants and Engineers, New Delhi, 2021 April.

Child D. (1990). The Essentials of Factor Analysis, Cassell Educational Limited, London.

Chokshi C.K. (1978). Construction contracts: disputes, causes and remedies. Proceedings of the Seminar on Construction Contracts, Disputes and Arbitration. Organized by Builders Association of India at Patna, India on 26 Nov.

Dennis D. (1991).Watershed in contractual history. Handbook and List of Members, The Chartered Institute of Building, London, 104-106.

Dillon W.R. and Goldstein M. (1984). Multivariate Analysis, John Wiley and Sons, New York.

Fox G.A. (1974) Subsurface construction contracts - a contractor's view. J. Constr. Div., ASCE, 100(2), 153-158.

Gau G.W.(1978). A taxonomic model for the risk-rating of residential mortgages. J. of Business, 51(4), 687- 706.

Hair J.F., Anderson R.E. and Tatham R.L. (1990). Multivariate Data Analysis, Macmillan Publishing Company, New York.

Ibbs C.W., Back W.E., Kirn JJ., Wall D.E., De La Garza J.M., Hassanein M.A., Schran S.M. and Twardock R.K. (1986). Determining the Impact of Various Construction Contract Types and Clauses on Project Performance. The University of Illinois at Urbana-Champaign, Report to The Construction Industry Institute, March 1986.

Ibbs C.W. and Ashley D.B. (1987). Impact of construction contract clauses. J. Constr. Engrg. and Mgmt. ASCE, 113(3), 501-517.

Iyer K.C., Kalidindi S.N. and Ganesh L.S. (2001). Effects of construction contract clauses on occurrence of disputes in India. Asia Pacific Building and Construction Management Journal, Hong Kong, 6(1), 13-24.

Jergeas G.F. and Hartman F.T. (1994). Contractors' construction-claims avoidance. J. of Constr. Engrg. and Mgmt., ASCE, 120(3), 553-560.

Maloney W.F. and McFillen, J.M. (1995). Job characteristics: unionnonunion differences. J. Constr. Engrg. and Mgmt., ASCE, 121 (1), 43-54.

Military Engineer Services (1989). General Conditions of Contract: IAFW - 2249. Government of India Press, Howrah.

Levin R.I.(1988). Statistics for Management, Prentice Hall of India Private Limited, New Delhi.

Overall J.E. and Klett C.J.(1972). Applied Multivariate Analysis, McGraw Hill Book Company, Inc., New York.

Rau A.N. (1993). Construction in eighth plan: approach of the Planning Commission. Souvenir of the XV all India Builders' Convention, Coimbatore, September 5-7, 57-61.

Roy S.R.K. (1991). Arbitration relating to construction contracts. Seminar on Contract Management & Arbitration in Civil Engineering Projects, The Institution of Engineers (I), New Delhi, 22-23 September, IV-21 - IV-25.

Russel J.S. (1990). Surety bonding and owner-contractor prequalification: comparison. J. Prof. Issues in Engrg., ASCE, 116(4) , 360-374.

Saluja V.K. (1984) Management of construction contracts. all India Seminar on Construction Contracts - Contribution by Contractor, Architects, Consultants and Engineers, New Delhi, 20-21 April.

SAS Institute Inc. (1988). SAS/STAT® User's Guide, Release 60 .0 03 Edition. Cary, NC: SAS Institute Inc. 1028.

Steen R.H. (1994). Five steps to resolving construction disputes without litigation. J. Mgmt. in Engrg., ASCE, 10(1), 19-21.

Thomas H.R., Smith G.R., and Wright, D.E. (1990). Resolving disputes over contract notice requirements. J. Constr. Engrg. and Mgmt., ASCE, 116 (4), 738-755.

Thomas H.R., Smith G.R. and Wright D.E. (1991). Legal aspects of oral change orders. J. Constr. Engrg. and Mgmt., ASCE, 117 (1), 148-162.

Thomas H.R., Smith G.R. and Wirsching S.M. (1995 a). Understanding defective specifications. J. Constr. Engrg. and Mgmt., ASCE, 121 (1), 55-64.

Thomas H.R., Smith G.R. and Cummings DJ. (1995 b). Have I reached substantial competion? J. Constr. Engrg. and Mgmt., ASCE, 121(1), 121-129.

Thompson LJ. and Portis C.T. (1978). History of evasive contract phrases. J. Constr. Engrg. and Mgmt., ASCE, 104(1), 525-537.

K. Chandrashekhar Iyer

Associate Professor

Department of Civil Engineering

Indian Institute of Technology, Delhi - 110016 (India)

Satyanarayana N. Kalidindi

Associate Professor,

Department of Civil Engineering,

Indian Institute of Technology, Madras - 600036 (India)

L. S. Ganesh

Professor

Industrial Engineering and Management Division

Indian Institute of Technology, Madras - 600036 (India)

Copyright Global Institute of Flexible Systems Management (GIFT) Dec 2002
Provided by ProQuest Information and Learning Company. All rights Reserved