Introduction
Cost accounting is an integral part of estimating a new construction project and estimating a project’s cost is the most important step of all. Estimating project costs is even mentioned in the Holy Bible: “Suppose one of you wants to build a tower. Don’t you sit down and estimate the cost, to see if you have enough capital to complete it? If you don’t, then when you have laid the foundation but can’t finish, all the onlookers start making fun of you and say, `This is the man who began to build, but couldn’t finish! ‘ Luke 14:28-30 (Sterm, 1998)
The State of Texas has some of the best highways in the nation. The Texas Department of Transportation (TXDOT) ensures this by spending roughly $6 to $7 billion dollars annually on new construction and maintenance projects throughout the state (Texas Department of Transportation, 2009). Most of this money is paid to outside contractors who bid on the jobs they would like to perform. Every contractors bid is an estimate of what they are willing to do a particular job for. Every contractor has its own way of estimating and that is the area I would like to focus on.
I would like to determine the common cost accounting (estimating) procedures used by TXDOT contractors and how they could be used to setup a better, more uniform cost accounting (estimating) system? If a uniform system is used the bids will be more competitive and TXDOT could save millions of dollars every year. At the end of fiscal year 2009 TXDOT had about $10 million dollars set aside for possible claims that contractors were placing because they had not estimated or bid properly. There was one case in 3 009 where a contractor withdrew their winning bid after they had been awarded a contract because of an estimating error on their part which led to major delays of the project (Wear, 2009). These claims and delays caused by estimating errors have a significant effect on TXDOT’s financial statements as well as its ability to gain future State and Federal appropriations. TXDOT has many manuals for its employees that tell them how to plan estimates and input them into their system but has very few actually estimating instructions for the contractors who do the bulk of their work (Texas Department of Transportation).
TXDOT estimates are composed of many different bid items, or items that are to be completed for the estimate. Each bid item is assigned a number that represents a certain category of work to be performed by the contractor. A description of the work to be accomplished under a bid item is available in the Texas Department of Transportation’s (TxDOT) Standard Specifications for Construction and Maintenance of Highways, Streets, and Bridges (otherwise known as the Spec Book) for standard specification items (Mark A. Marek, 2009).
This Spec Book is virtually the only thing that a contractor has to go by when planning an estimate and it was lasted updated in 2004 (Texas Department of Transportation, 2004). It does not mention anything about estimation only what is involved in and how to perform each item. It is my belief that if a uniform cost accounting (estimation) plan was put into effect for all contractors doing work for TXDOT then TXDOT would save a lot of time and money trying to sort out problems with contractors estimates. With my research I 4 ill work to find the common cost accounting (estimating) procedures used by the top contractors that do work for TXDOT and determine if a uniform system is possible. If one is not possible at least we may have an idea of where some of these companies are going wrong and where improvements can be made within the TXDOT system to help prevent them from happening in other contracts. Literature Review Finding information directly related to TXDOT cost accounting was a difficult tasks as I found no studies of TXDOT in my searches so far.
The literature I did find has all dealt with the cost accounting and estimating and ways people have sought to improve them. Jack A. Lazarczyk discussed why Engineers and Accountants should work together to provide a greater assurance during an audit of a construction project. Accountants know and understand the financial side where as the engineers know and understand the issues of scopes, methods, and quality. This approach could be used in the estimating of future projects as well. Accountants and engineers should work together from the beginning of the roject until after the project is complete. This would help to ensure that all financial numbers are in accordance with accounting and engineering standards (Lazarczyk, 2009). Randall S. Wyandt proposed the idea that a third party should handle the project controls and costs as a project manager might want to “fudge” the 5 numbers to give a more optimistic outlook. Also states that to complete a project close to budget the original budget needs to be accurate. An accurate budget is only as good as the cost estimating procedures it is backed by.
If the original estimate or bid for a job is fudged even the tiniest bit in order to make sure that a contractor wins the job then the job is already off to a horrible start (Wyandt, 2009). Kul B. Uppal stated that cost methodology is another important part of total cost management. That methodology should consist of standard set of cost estimating practices, resources and tools that are consistently applied in all cost estimates. Methodology is directly related to the development of cost controls techniques and tools along with measurement of cost levels and trends.
Consistency is key to ensuring that TXDOT is getting the most accurate and best bid (estimate) possible for each and every one of their contracts (Uppal, 2009). Robert H Harbuck discussed Life Cycle Cost Analysis (LCCA) which is an economic assessment technique that is used to evaluate competing infrastructure alternatives based on the estimated total cost of ownership. The use of LCCA gained prominence in the 1960s in response to US government agencies desire to improve the cost effectiveness of equipment procurement. Its use is common in some industries but its record in the area of transportation is sporadic at best (Harbuck, 2009 ).
Alexia A. Nalewaik and Jeffrey Witt looked at the challenges of reporting project costs and risks by outlining the differences between cost accounting 6 and cost (project) control. They also looked at some of the pitfalls of cost accounting systems and of project reporting and how they can better be used when combined. Project reports should fully show the current standing of the project. It should also show how the project is doing versus its original cost estimate. This will help the company who is performing the work to better estimate the next job they bid on (Nalewaik & Witt, 2009). Steven M.
Trost and Garold D. Oberlender contemplated the ability of predicting the accuracy of early cost estimates using factor analysis and multivariate regression. Information was collected from 67 competed projects; this information came from 22 companies, consisting of 16 owners and 6 contractors and engineering firms. The data included estimated costs, actual costs, and a rating system based on 45 variables, or elements, deemed potentially important to the accuracy of early estimates. They rated each element on a one-to-five Likert scale, with “one” representing the best score and “five” representing the worst score.
They split the 45 factors into 11 factor groups and performed a multivariate regression analysis on the factor scores of these 11 factors. They concluded that the most significant factor was basic process design. The second factor of significance was team experience and cost information. Time allowed to prepare the estimate ranked third among the significant factors influencing estimate accuracy (Trost & Oberlender, Predicting Accuracy of Early Cost Estimates Using Factor Analysis and Multivariate Regression, 2003). 7 Dr. Murat Ciraci and Dr.
Deniz Ayse Polat researched the importance of accurate estimates at the early stages of capital projects. They realized how important early cost estimates are to the owners and contractors. Inaccurate early estimates can lead to lost opportunities, wasted development effort and lower than expected returns. An accurate cost estimate, made at the planning stage, has the potential to avert many problems. Using the information from Trost and Oberlender they discovered that the factors that influence the cost estimate accuracy the most were ones that dealt with the project’s scope.
They conducted an evaluation of 39 different cost estimate methods developed for usage at the planning stage. The evaluation looked at the cost factors they employ and the extent to which they take the project scope into consideration. The authors also surveyed 13 published articles dealing with the factors that influence cost estimate accuracy. When compared with their first evaluation it became clear that project scope as well, as many other relevant factors, were not receiving the appropriate consideration by many of the estimate methods (Ciraci & Polat, 2009).
Kevin McDowell suggested that not only present data collection but also historical project information is needed for future project estimation. His research was primarily in the oil and gas industry but the industry estimation strategies and procedures could easily be used in the transportation construction industry. Determined 5 key elements needed to enhance how information is used: 8 ? Project information- certain information is needed to fully understand the project. ? Cost elements- Common cost elements allow for data to be easily grouped and compared. Code of accounts- Improves the ability to collect historical costs and provides an effective basis for project control systems. ? Data collection and capture- How the data is collected greatly affects the reliability of it. ? Reporting of cost metrics-Provide a guide with which to see if the estimate is aligned with expected results (McDowell, 2009) Martinez, Migliaccio, Zandbergen and Zhang tested the validity of the nearest available interpolation method used by the cost estimation classification system established by the Association for the Advancement of Cost Engineering International (AACE International).
They determined that the method was defensible due to strong spatial autocorrelation at both the national and state levels. They proved that that the proximity-based interpolation for location adjustment is valid, but to what extent. They suggest that future studies should be performed to further test their results against other theories (Martinez, Migliaccio, Zandbergen, & Zhang, 2009). In a report to the Subcommittee on Energy and Water Department, Committee on Appropriations, House of Representatives the Government 9 Accountability Office stated that a high-quality cost estimate is credible, well-documented, accurate, and comprehensive. Credible- when it has been cross-checked with independent cost estimates, the level of confidence associated with the point estimate has been identified, and a sensitivity analysis has been conducted—that is, the project has examined the effect of changing one assumption related to each project activity while holding all other variables constant in order to identify which variable most affects the cost estimate; ? Well-documented- when supporting documentation is accompanied by a narrative explaining the process, sources, and methods used to create the estimate and contains the underlying data used to develop the estimate; Accurate- when it is not overly conservative or too optimistic and based on an assessment of the costs most likely to be incurred; and ? Comprehensive- when it accounts for all possible costs associated with a project, is structured in sufficient detail to ensure that costs are neither omitted nor double-counted, and the estimating teams’ composition is commensurate with the assignment (United States Government Accountability Office, 2010). MethodologyData Collection I will be following the methodology that Oberlender and Trost used in their 2001 and 2003 studies.
Their studies led to the creation of an estimate grading software that I believe may be beneficial to TXDOT. The difference between their study and mine will be that I will strictly look at transportation projects where theirs looked at construction estimates from multiple markets. I will be using the 10 top 20 transportation contractors in Texas as ranked by Texas Construction Magazine at the end of 2009 as my participants. They are ranked by the amount of work they performed in 2009 in millions of dollars. These transportation contractors did a combined $2. billion dollars worth of work in Texas during 2009.
The 45 elements were grouped into four 11 divisions. Division I represents who was involved in preparing the estimate, Division 2 refers to how the estimate was prepared, Division 3 involves what was known about the project, and Division 4 represents other factors affecting the estimate. Fig. I (Appendix) shows the 45 elements of the estimating scoring system. A rating scale for each element was determined, ranging from one to five, with one as “best” and five as “worst. The one-to-five rating scale conforms to a Likert scale and does not force the respondent to rate a particular element “good” or “bad” but allows a neutral response. The team also developed descriptions for each of the 45 elements. Since a certain amount of subjectivity would still be involved in rating the elements (e. g. , a rating of “two” to one person may not be same as a “two” to someone else), the research team developed suggested ratings for each of the possible responses for each of the elements. A two-page questionnaire was compiled that contained the elements to be rated on one side (Appendix Fig.I) and the requested specific project and cost data (Appendix Fig. 2) on the reverse side. Cost data will be requested as both estimated and actual costs. As a part of the estimated cost data, the respondents will be asked to list the amount of contingency applied to the estimate.
The contingency amount, broken out separately, will enable comparisons between the base estimate and the actual costs. Additional cost data will be requested including bulk materials, engineered equipment, engineering design and construction. 2 Oberlender and Trost’s team contemplated that in order for this study to be valid and reliable that they would need to collect at least 20 observations more than the number of variables; with 45 variables, 65 scores will be needed (Trost & Oberlender, Predicting Accuracy of Early Cost Estimates Based on Estimate Quality, 2001). I will be asking each of my 20 participating contractors to provide cost data for 3 to 5 of their current or completed construction projects. This will give me between 60 to 100 responses.
I will also gather information from TXDOT concerning what was used to produce their estimates for the projects that I get a response from so that I can see what differences lie between the State’s initial estimate and each contractor’s initial estimate. MethodologyData Analysis Oberlender and Trost’s initial data analysis involved a multivariate regression on the 45 elements, based on their 67 observations. The analysis produced varying results depending upon which of the 45 elements were included in the analysis and in which order they were added or removed during the analysis.
Because all the elements related to the issue of estimate accuracy, and many were related to each other, multicollinearity was a distinct problem (Trost & Oberlender, Predicting Accuracy of Early Cost Estimates Based on Estimate Quality, 2001). A factor analysis was performed on the data to overcome the multicollinearity problem. This analysis provides a deterministic method to group variables into meaningful subdivisions the number or which can be anywhere 13 between one and the total number of variables.
For this research, the number of factor groups could range from one to forty-five (Trost & Oberlender, Predicting Accuracy of Early Cost Estimates Based on Estimate Quality, 2001). Several guidelines are available to assist statisticians in determining how many factors to include in a factor analysis. The researchers utilized the minimum eigenvalue criterion as a starting point to determine the proper number of factor groups in the final analysis. An 11 factor model was chosen, because the resultant groups made the most theoretical and practical sense. Table 1(See Appendix) shows the 45 elements grouped into the 11 factor model.
The nature of factor analysis creates orthogonal factors and solves any multicollinearity problems (Trost & Oberlender, Predicting Accuracy of Early Cost Estimates Based on Estimate Quality, 2001). Since Oberlender and Trost’s team have already discovered the multicollinearity I will use their 11 factor groups to conduct a multivariate regression on the data collected from my participants. The goal is to find the same results as Oberlender and Trost. If my results differ from theirs I will perform my own factor analysis and redo my multivariate regression analysis using the new factors.
I will then take a ratio of the amount of claims (adjustments to the original estimate) that each individual contractor placed on their respective jobs to the total amount of work they performed on those jobs. I am hoping to find that the contractors with the lowest percentages of claims will also be the ones with 14 good cost accounting/estimating procedures in place. I will use the ranking system of the methods they use to see which methods are best for estimating and/ or accounting for costs on the jobs and develop a list that I believe will help all contractors better estimate their jobs.
I will also look into the software developed from Oberlender and Trost’s research and determine if it is something that could help the individual contractors with preparing the best bid or if it would help TXDOT to choose the best bid instead of the lowest bid. Expected Outcome I expect to find that there are several common estimating practices used by the contractors of TXDOT and that these estimating practices can be ranked and used to create a uniform cost estimating/accounting procedure manual to be used by all TXDOT contractors.
I believe this would help to save TXDOT millions, if not billions, of dollars everywhere. This could lead to an increase in the number of projects they are able to perform each year or allow the State and Federal Governments to allocate those savings to more important causes such as Education.
March 15, 2011 Receive approval to continure with study April 15, 2011 Make contact with contractors and mail off questionnaires June 1, 2011 Deadline to receive questionnaires August 1, 2011 Completed Data Analysis September 1, 2011 Final Report Issued 19
References
- Ciraci, D. M. , & Polat, D. D. (2009). Accuracy Levels of Early Cost Estimates, in Light of Estimate Aims. Cost Engineering , 16-24. Harbuck, R. H. (2009 ). Life Cycle Cost Analysis for Transportation Projects. AACE International Projects , RISK. 05. 1-RISK. 05. 13. Lazarczyk, J. A. (2009).
- Project Audit- Joint Engineer and Accountant Team Approach. AACE Intenational Transactions , OWN. 05. 1-OWN. 05. 7. Mark A. Marek, P. (2009, May 1). PS&E Preparation Manual . Texas. Martinez, A. A. , Migliaccio, D. G. , Zandbergen, D. P. , & Zhang, S. (2009).
- Assessment of Methods for Adjusting Construction Cost Estimates by Geographical Location. AACE Internationa Transactions , EST. 15. 1-EST. 15. 12. McDowell, K. (2009).
- Effective Project History Collection and Retrieval- Back to Basics. AACE International Transactions , EST. 07. 1-EST. 07. 6. Nalewaik, A. A. , & Witt, J. (2009).
- Challenges to Reporting Project Costs and Risks to Owner Decisionmakers. AACE International Transactions , OWN. SO2. 1-OWN. S02. 10. Sterm, D. H. (1998).
- Complete Jewish Bible. Jewish New Testament Publications, Inc. Texas Department of Transportation. (n. d. ). Online Manuals. Retrieved October 31, 2010, from TXDOT: http://onlinemanuals. txdot. gov/manuals/index. html Texas Department of Transportation. (2009).
Operating Budget for Fiscal Year 2010. Austin. Texas Department of Transportation. (2004, June 1).
Standard Specifications for Construction and Maintenance of Highways, Streets, and Bridges.
- Texas. Top Contractor List Breakout Rankings – Building/Market Sectors. (2010). Retrieved Ocober 30th, 2010, from Texas Construction: http://texas. construction. com/texas_top_lists_construction_firms/2010/topcontractors/market/00 1-100. asp Trost, S. M. , & Oberlender, G. D. (2001).
- Predicting Accuracy of Early Cost Estimates Based on Estimate Quality. Journal of Construction Engineering and Management , 173-182. Trost, S. M. , & Oberlender, G. D. (2003).
- Predicting Accuracy of Early Cost Estimates Using Factor Analysis and Multivariate Regression. Journal of Construction Engineering and Management , 198-204. United States Government Accountability Office. (2010).
- Actions Needed to Develop High- Quality Cost Estimates for Construction and Environmental Cleanup Projects. 20 Uppal, K. B. (2009).
- Cost Estimating, Project Performance and Life Cycle. AACE International Transactions , TCM. 03. 1-TCM. 03. 9. Wear, B. (2009, November 19).
- I-35/Ben White flyover project delayed. The Statesman . Austin, Texas. Wyandt, R. S. (2009). A Case for Third Party Project Controls. AACE International Transactions , OWN. 06. 1-OWN. 06. 9.
