Structural Equation Models and Two Stage Least Squares Models, Research Paper Example
Words: 2519Research Paper
This is an empirical study of the of the an order-by- mail company’s model for the month on the total amount of the offer that is purchased based on the total amount of book purchases and CD purchases that company make and the type of offer that the company gives to its club members. To help carry out this empirical study, there is the use of the ordinary least squares and the Heckman two-stage regressions for the possible endogeneity of the company’s adoption decision. The outcome of the analysis show that there is a correction between the money spent on books and CD’s and money spent on special offers hence existence of a feedback loop between the two predictors and the response.
Introduction/ purpose statement
The most important basics price modeling offered in the econometric modeling literature indicates that there has existed a direct relationship between the prices charged on goods on offer and the quantity of the goods sold. The price model is also known to positively impact on the outcome of the sales made in terms of the volumes of goods sold. There has however been limited empirical evidence on this model despite several studies undertaken in the same especially on the same nature of business that have offers. Banker et al. (1993) and Srinivasan (2000) talks about the same
Literature and hypothesis development
Literature Review: changing the pricing Model
The order-by- mail company literature yields contradictory results concerning the effectiveness of the company on firm level performance. In contrast, extant firm–level studies indicate that firm performance is positively related to company’s adoption and/or to the ‘competitiveness’’ of the firm. What remains unclear from this literature is whether performance incentives mediate the relation between firm ‘‘Competitiveness’’ and firm performance, as would be predicted by the standard new pricing model.
There are a few empirical studies that relate firm adoption and/or firm ‘‘Competitiveness’’ to underlying incentives. Banker et al. (1993) show that the feedback of service performance outcomes to line workers is positively associated with the implementation of model at the firm level, which suggests that incentives motivate actions. However, this study does not investigate firm performance outcomes, so it provides no information about whether incentive-driven actions affect performance. In a field study of a division of a single service firm, Young and Selto (1993) try to relate workgroup (employee) ratings to company’s implementation/performance and incentives. Unfortunately, the division manager refused to let participants respond to survey questions regarding incentives. Further, Young and Selto (1993) found that personnel ratings and compensation were unrelated to worker performance or to company’s implementation or performance.
In a longitudinal field study of a single service firm, Chenhall and Langfield- Smith (3003) report on the impact of incentive changes, such as a gain-sharing reward system, and company/ total quality management (TQM) adoption on performance outcomes. Only minimal reference is made to the actions engendered by the incentive changes and company/ TQM adoption. The authors find that, although gain sharing and company/TQM adoption led to increased productivity during the initial decade, additional organizational changes in subsequent years, such as team-based structures and value added management (VAM), were relatively unsuccessful.
The authors attributed the lack of success to reduction of trust between workers and management engendered by the mechanistic control of gain sharing and the intrusiveness of VAM monitoring.
In a more comprehensive study, Sim and Killough (1998) investigate inter alia the relationships (1) among firm ‘‘ORDER-BY-MAIL Competitiveness,’’ customer satisfaction performance goals, and customer contingent incentive rewards, and (3) among firm ‘‘company’s Competitiveness,’’ quality performance goals, and quality-contingent incentive rewards for firms in the electronics industry.
Counterintuitively, they find that customer satisfaction performance in firms decreases significantly with customer-related performance goals and is unrelated to customer-related incentive rewards. However, they do find that customer satisfaction performance increases significantly with both the interaction between firm ‘‘ORDER-BY-MAIL Competitiveness’’ and customer-related performance goals and with the interaction between firm ‘‘ORDER-BY-MAIL Competitiveness’’ and customer-related incentive rewards. Surprisingly, they also find that quality performance is unrelated to quality-related performance goals, quality-related incentive rewards, or the interaction of ‘‘company’s Competitiveness’’ and quality performance goals. They further show that quality performance of the company
Data and summary statistics
Table 1 Panel A shows that firms are significantly larger in terms of sales revenue and more profitable than firms. These results raise potential causality concerns. Is it that firms are more profitable, or are profitable firms more likely to adopt the model? The Incentive Index quantifies the response by firm managers to the following survey question with respect to six firm performance measures: ‘‘How important are the following measures in evaluating the performance of your service system: inventory turns, equipment utilization, labor utilization, on-time delivery, scrap/waste reduction, and quality?’’
Summary Statistics and Univariate Tests
Panel A: Production Data in Company and Industry
Impl .model All Firms Mean Industry Mean Firms Mean t-statistic Mann-Whitney
(SD) (Median) (Median) (Median) (P-value) (P-value)
Sales 25.9 23.1 38.5 _1.76 _3.44
Revenue (18.0) (15.6) (35.6) (0.08) (0.03)
Production 17.3 17.0 30.6 _1.38 _3.41
Costs (13.9) (11.8) (18.1) (0.17) (0.03)
Operating 6.0 5 .0 7.8 _3.33 _3.38
Profits (6.3) (3.7) (7.0) (0.03 (0.03)
Scores are based on a five-point Likert scale where 5 extreme importance and 1 no importance. Table 1, Panel B provides means and medians for the Incentive Index and each of its components stratified by company and non-company subsamples, and also by high- and low-profitability subsamples. Means and medians are similar to each other. Overall, the Incentive Index is significantly larger for firms than for non-firms based both on a t-test and a Mann-Whitney test, suggesting that firm managers in firms are subject to greater production incentives than in non-firms. Looking at the separate incentive components, as expected, firms consider inventory turns and scrap/waste to be significantly more important in evaluating their service system than non-firms.9 More surprisingly, the four other incentive component differences are insignificant, but this may be due to the difficulty of measuring such variables as equipment utilization, labor utilization, and quality when compared with inventory turns and scrap/waste.
Panel B: Implementing model in the firm and by High/Low Operating Income/Sales
Impl .model All Firms Mean Industry Mean Firms Mean t-statistic Mann-Whitney
(SD) (Median) (Median) (Median) (P-value) (P-value)
Incentive 24.13 25.98 _3.34 _3.42 24.9
Index (24.00) (27.00) (0.00) (0.00) (25.50)
Inventory 3.00 3.50 _3.00 _3.06 3.46
Turns (3.00) (4.00) (0.00) (0.00) (3.50)
Equipment 4.45 4.48 _0.23 0.02 4.30
Utilization (5.00) (5.00) (0.82) (0.99) (4.50)
Labor 4.50 4.47 0.22 _0.12 4.38
Utilization (5.00) (5.00) (0.82) (0.91) (5.00)
On-Time 4.67 4.78 _0.96 _1.23 4.69
Delivery (5.00) (5.00) (0.34) (0.22) (5.00)
Quality 4.65 4.78 _1.43 _1.46 4.78
In view of the fact that performance as measured by operating profits, the contribution margin, and total production costs yield qualitatively similar results, we provide tables for operating profits only. In addition, we show results for the Outcome Index performance measure because the latter metric yields results that are at variance somewhat with those of operating profits and the other performance measures. Using ordinary least squares (OLS), firm operating profits normalized by sales revenue is regressed on the Index, the Incentive Index, and on a number of control variables. If profitability is related to actions (conditioned on incentives), then the coefficient on the Index will be positive and significant. If performance is related to incentives (conditioned on actions), then the coefficient on the Incentive Index will be positive and significant.
Table 2, Panel A shows results for the case where the regression does not include an interaction term between the two indices. We further test H1 in Panel B of Table 2 by incorporating an interaction term between the Incentive Index and the Index. To mitigate the issue of multicollinearity in the presence of interaction terms and to enhance interpretability of the regression, we de-mean all non-dummy regressors (Aiken and West 1991) in the regressions that follow.
The regression in Table 2, Panel A yields a statistically significant F-statistic (F _ 37.65, p _ .000) with an adjusted R3 of 73 percent. The Index is positive and significant (t _ 3.41, p _ .001), indicating that firm performance is positively associated with ‘‘ORDER-BY-MAIL Competitiveness.’’ The Incentive Index is not significant (t _ 0.58, p _ .567), indicating that firm performance is independent of firm managers’ incentives, a result that is not supportive of H1a.13 Consistent with H3, the experience variable is positive and highly significant (p _ .000), indicating that firms that have more experience with are more profitable. We further find that firms that belong to firms that are international in scope are significantly more profitable (t _ 3.54, p _ .001). Also, firms for which adoption require the firm to increase financing are less profitable, but the coefficient is only marginally significant at the one-tailed level (t _ _1.41, p _ .166). Finally, automotive parts firms are significantly less profitable than electronics components firms (t _ _4.43, p _ .000). The Table 3, Panel B regression incorporates an interaction term between the Index and the Incentive Index but is otherwise identical to Panel A. The interaction term control
Two- stage regression analysis
The following window was generated after the execution of the TSLS parameter which give the estimates that prettily close to the true values of the output.
This is the most likely window that sis produced after same process has been repeated severally with different value.
To make sure there is stability and of results in the results of this study, several robustness checks are cared out. The levels of significance is for instance not based on the Heckman standard error but on bootstrap and jackknife standard errors which still give the desired qualitative results for all the regression analysis carried out in the cause of this study. Even in the cases where more than one firms are owed by the same company. This accounts for companies that have potentially correlated data. This however is not possible with their regressions re-estimated after the dropping has been made. The results in all the cases were however not affected by this procedure since the company employs inventories that are material requirements planning in nature (MRP) which indicate that the benefits of the company are ascribed to this model. There was however the dropping of some of the models that were initially used by the company though it had no adverse effects on the company results on the efficiency of the model.
Based on the data set that is the study uses, has make several contribution to the empirical price modeling and the literature of such companies that operate the business of mail-by-order and have offers. The study show that based on the empirical results of the study, businesses is encouraged to adopt this kind of business model but not the specific practices of the model. According to this study it is evident that the essential inventory issues like the inventory turns are some of the motivational aspects for the adoption of the model. There is however need for more research on other aspects of the model like its quality to make the model more clear.
The measure of the performance of the profitability of the model and the experiences of using the model indicates that there is significantly related to the performance as measured by the outcome index.
There are however several reasons based on the findings that would have caused the company to adopt to the new system of pricing and the model more the specific incentives and practices that would have caused the company to invest in the model. Not only do the mere performance goals of the company would not have been the only motivating factor that would have made the company implement the model part from the underlying monetary reward that is attached to the model. This is in tandem with the career incentives that would have been argued out by Holmstrom (1983) and Dewatripont et al. (1999) generalizations. This makes the use of the company study findings be used fairly intensively and widely. The variables found in Heckman analysis in this case may have played a major role in the management decision to adopt the model though its OLS regression commonly known as the second-stage Heckman regression. Some of the incentives that would however motivated the managers of the company to adopt this model might not have been found or observed at the cause of the study yet ar truly motivators behind the practice decisions.
There are several weaknesses that the study but three of these weaknesses are found to be the most pertinent and found out to have affected and greatly influenced the results of the study. The size of the sample is one of these weaknesses as the sample had to be gets smaller, the accuracy of the study also reduces hence need to deal with larger samples of the data in order to improve on the accuracy and acceptability of the findings of the study. A larger sample would also allow for an analysis of the endogeneity that is more robust. This is possible through the use of the parametric and non parametric techniques. This is however difficult has it as it is not easy to have data that is not proprietary in nature even if it means working with smaller samples.
To avoid the several errors that are common with measurement, the data which is cross-sectional in nature is desideratum in nature get care has to be taken to mitigate the error. Incorporating the performance data for the company may give more room for a more insightful analysis of the endogeneity of counterfactuals through treatment effects that are associated with those estimated by the company. This means that the turns yield as a result of this model is higher as shown by the empirical results of the model.
Ahmed, N. U., E. A. Runc, and R. V. Montagno. 1991. A comparative study of U.S. manufacturing firms at various stages of just-in-time implementation. International Journal of Production Research 29: 787–802.
Aiken, L. S., and S. G. West. 1991. Multiple Regressions. Beverly Hills, CA: Sage Publishing.
Balakrishnan, R., T. J. Linsmeier, and M. Venkatachalam. 1996. Financial benefits from JIT adoption: Effects of customer concentration and cost structure. The Accounting Review 71: 183–205.
Banker, R. D., G. Potter, and R. G. Schroeder. 1993. Reporting manufacturing performance measures to workers: An empirical study. Journal of Management Accounting Research 5: 33–55. ,
Lee, and G. Potter. 1996. A field study of the impact of a performance-based incentive plan. Journal of Accounting and Economics 21: 195–226.
Time is precious
don’t waste it!