<p>This paper reconsiders the information content of losses, specifically, the extent to which losses contain distinct and offsetting information about future cash flows and about risk. Based on theory that suggests exit value is the lower bound of firm value, I posit that shareholders who decide to abandon the firm (or some portion thereof) will receive the exit value of disposed resources, thereby resolving uncertainty about payoffs (that is, cash flow uncertainty). Under this view, a higher likelihood of abandonment, proxied by losses, should be associated with both lower payoffs (the exit value of the disposed net assets) and lower risk (because uncertainty about the payoff is partially resolved). Using Vuolteenaho's (2002) method to decompose realized returns into expected returns, cash flow news and discount rate (risk) news, I predict and find that losses provide adverse news about cash flows (the valuation numerator) and favorable news about the discount rate (the valuation denominator). Because the effects of the two types of news are mutually offsetting, the relation between earnings and returns appears weaker for loss firms than for firms reporting profits. These results suggest that losses are valuation relevant, in the sense of providing information that is correlated with the information in returns.</p> / Dissertation

This study examines whether institutional investor shareholdings inhibit firm managers from engaging in earnings management practice. It investigates the empirical association between discretion/flexibility available to managers in managing abnormal non-cash working capital accruals and institutional stock ownership for a sample of 386 New York Stock Exchange firms over a period of 8 years, from 1991 through 1998. The differential institutional influence on the level of accrual management of firms having different information environment, S&P 500 versus non S&P 500, is also examined to see whether the difference in information environment of these two sets of firms has any effect on this empirical relationship. By performing various multivariate statistical analyses, I find significant evidence that institutional stockholders reduce management flexibility in generating abnormal accounting accruals. Further, concentrated institutional shareholdings in some cases are found to diminish managerial propensity to manage abnormal accruals. A separate analysis for the S&P 500 and the non S&P 500 firms reveals that institutional monitoring effect on accrual management is different for these two sets of firms. I observe that institutions do not have mitigating influence in the S&P 500 firms but have significant mitigating effects on accrual management level in the non S&P 500 firms. The study makes two-fold contributions to the existing earnings management literature. First, it is generally assumed in prior studies that firms have uniform abilities to generate abnormal accruals to manage earnings. This study provides evidence that managements ability to manage earnings is not constant across firms but varies according to the level and concentration of institutional stock ownership. Institutional investors are found to improve the quality of corporate governance in financial reporting in cases where other important governance factors exist. Consequently, this study also extends prior research that examined the effects of other influential governance factors such as external audit, independence of boards or audit committees on the level of accrual management. Second, I develop a unique and powerful accrual model, which represents an improvement over the traditional models typically used in previous research and provides more robustness to the tests of earnings management.

Unique to survival analysis of veterinary clinical data is classification of observations from euthanized animals. The first study highlighted limitations of Kaplan-Meier product limit analysis (KM) of veterinary clinical data. Three data sets with different outcome proportions (alive, lost-to-follow-up, dead due to disease, dead due to other, euthanized due to disease, euthanized due to other) were used. Different classifications of observations from euthanized animals caused inconsistent conclusions of significant differences between strata within data sets. At times, ranking of median survival time estimates for strata was reversed. The KM was found inappropriate to evaluate observations from euthanized animals. This finding, coupled with restriction of KM to two-state description of disease (alive to outcome), prompted exploration of an alternate analysis method. Markov models allow modeling of multiple health states and outcomes. A 5-state, time-homogeneous, Markov chain was used for a cohort of 64 dogs with generalized lymphoma. The model contained two transient (WELL, TOXIC) and three absorbing (DEAD, EUTHANASIA, LOST-TO-FOLLOWUP) states. The transition probability matrix (P) was used to iterate future transitions and survival probabilities. Matrix solution and Monte Carlo simulation were used to estimate survival time. Estimates appeared reliable. Markov modeling was extended for comparison of vaccine-associated sarcoma progression after treatment in a cohort of 294 cats. For a 5-state model, transition probabilities derived from exponential transformation of incidence rates were used to construct P for each treatment - NONE (no surgery), SX (surgery) and SX+RAD (surgery and radiation). Monte Carlo estimates of durations in transient states and expected survival showed SX+RAD prolonged expected survival significantly longer than SX than NONE. Commitment to repeated treatment with surgery and radiation did prolong expected survival of cats with vaccine-associated sarcoma. Assumptions of Markov modeling did not appear prohibitive for analysis of veterinary clinical data and further exploration is warranted.

New varieties are provided to the Louisiana sugarcane industry by researchers at Louisiana State University AgCenter, the United States Department of Agriculture-ARS, and the American Sugar Cane League of the USA, Inc. Currently, Louisiana farmers plant sugarcane at rates ranging from two to five or more whole stalks. A two-stalk planting rate is used to plant all stages of the LSU AgCenters sugarcane variety development program. The objective of this study was to determine the effect of planting rate on sugarcane variety trial data and interpretation. A planting rate by variety experiment was conducted at the LSU AgCenters Sugar Research Station. A randomized complete block design was used for the experiment with three replications, eight clones, and three planting rates consisting of two, three, and four whole stalks. Increasing planting rate from two to four stalks significantly increased sugar yield by 11 to 15%. Cane yield and stalk population significantly increased when planting rate increased from two stalks to either three or four stalks in the 2001 plant cane crop. Stalk population and mean stalk weight were negatively correlated, thus the lower stalk populations tended to compensate with greater mean stalk weight. Theoretical recoverable sugar was not significantly different regardless of planting rate. As expected, clones differed for sugar yield and its components. Of utmost importance, the planting rate by clone interaction was not significant for any trait in any of the experiments. Thus, increasing the planting rate from two stalks to three or four stalks did not change sugarcane variety / clone ranking. A germination study was also conducted in the Fall of 2002. A randomized complete block design was used in this experiment, consisting of eight sugarcane clones replicated three times. The planting rate was two stalks planted at two locations at the Sugar Research Station. Sugarcane variety / clone germination was similar for both soil types. The Pearson correlation coefficients indicated taller stalks also had more buds per stalk than did shorter stalks. Mean stalk weight was greater for taller varieties. Mean stalk weight tended to increase as the number of buds increased on a stalk.

Coastal erosion and wetland deterioration are serious and widespread problems affecting Louisianas coastal zone. The application of agronomic and molecular techniques for improving crop species is well documented. However, these have not been routinely applied to species of ecological and environmental value. Spartina alterniflora is used extensively for shoreline protection and tidal marsh restoration because of its aggressive spreading habit and tolerance to salinity. The progress of marsh revegetation projects is limited by the costs and labor associated with vegetative propagation of Spartina. Hence, a breeding program was initiated to develop improved smooth cordgrass accessions with superior seed producing ability to accelerate coastal restoration projects by developing a seed-based propagation. The objectives were to: 1) evaluate the variation among the accessions collected from South Louisiana S. alterniflora native populations, and characterize accessions selected for use in genetic improvement; and 2) assess variability at the molecular level among selected plants used to establish a breeding program. One hundred twenty-six accessions of S. alterniflora were collected across South Louisiana in 1998. The accessions were characterized for location, date of collection, seed weight, and percent germination. Biplot and cluster analysis were used to analyze patterns of variation among accessions and locations. Vegetative and reproductive traits were evaluated during the growing seasons. Significant differences occurred among accessions for traits measured during vegetative stage and days to first panicle emergence. Date of collection contributed to overall variation among accessions, reflecting differences in maturity at the time of collection. Vegetative growth and differences in rust reaction allowed characterization of S. alterniflora accessions. Positive correlations were observed among vegetative traits, and negative correlations between those traits and rust reaction. Seven superior genotypes were selected for future population improvement. To assess genetic diversity within and between superior accessions, molecular and phenotypic characterizations were used to compare 40 selected genotypes, which were subjected to DNA fingerprinting using RAPD and AFLP. Cluster analysis results revealed considerable natural variation among the original collections for traits contributing to plant establishment from seed. The differences in clustering pattern demonstrate the usefulness of molecular markers in assessing genetic diversity more accurately.

This paper examines empirically whether sophisticated speculative short sellers can detect earnings management by targeting stocks with large income-increasing discretionary accruals and high total accruals. Prior research indicates that total accruals are overpriced and this overpricing is largely attributable to the mispricing of discretionary accruals. Recent studies show that neither auditors nor financial analysts utilize information in accruals. Using samples of 11,537 firm-quarter observations and 5,118 firm-year observations for 1,146 12/31 non-financial NYSE firms from 1992 to 1999, I find supporting evidence those speculative short sellers can detect earnings management using financial accounting information disclosed in 10-Q and 10-K report. Specifically, I identify a significant and positive association between relative short interest and quarterly accruals. When I decompose accruals into its discretionary and non-discretionary components, I find that quarterly discretionary accruals are positively and significantly related to relative short interest. I further divide quarterly data into four sub-samples of separate fiscal quarters and find that speculative short sellers detect earnings management especially in the third and fourth quarters of a fiscal year and trade consistent with the information provided in quarterly accruals. In addition, the empirical results indicate that speculative short sellers establish short positions in firms with high accruals and large income-increasing discretionary accruals estimated using annual financial accounting information.

The purpose of this research project is to determine whether the implicit theory of integrity, a theory from the social psychology literature that predicts how social judgments and decisions are made, can explain internal auditors decisions. The implicit theory (Dweck and Leggett 1988) states that there are two types of people: (1) entity theorists and (2) incremental theorists. Entity theorists form strong inferences from observed behavior that are used to predict future behavior. Incremental theorists, on the other hand, do not infer characteristics from behavior, and therefore, do not attempt to predict future behavior. In an internal auditing context, the implicit theory is applicable to an internal auditors assessment of managements integrity. A quasi-experiment was used to assess the main effect of integrity cues (three levels of integrity between subjects) on an internal auditor's assessed risk of management fraud based on a hypothetical case. The auditors implicit theory of integrity (interacted with the integrity cue) is also investigated. These tests offer some evidence in support of the main effect of the integrity cue on the fraud risk assessment, but no evidence is found in support of the interaction effect of the implicit theory.

The purpose of this research endeavor is to investigate the effect that the degree of participative congruence has on the relationship between program commitment and managerial performance. The degree of participative congruence (Clinton and Hunton 2001) is a new measure of participation designed to determine the amount of congruence between the perceived need for participation and the degree of participation allowed. This study used a questionnaire to evaluate the antecedents of and effects of the degree of participative congruence (DPC) in a budgeting setting. In order to investigate these effects, the questionnaire was mailed to a sample of 1,500 AICPA members. The results of the SEM offer some support of the proposed model; however, the results provided stronger evidence in support of an alternative model. Program commitment has a significant relationship with DPC, but there is no significant relationship between DPC and managerial performance. Evidence suggests that DPC has a positive effect on performance through the positive effect that DPC has on program commitment.

The enactment of the 1998 Internal Revenue Service (IRS) Reform Act resulted directly from perceived lapses in the federal taxation administration, which was publicly highlighted in Congressional hearings. Congress reacted by fundamentally altering the IRSs implementation of the Internal Revenue Code (for the first time since 1952). The Joint Committee on Taxation noted that the overall objective of the 1998 IRS Reform Act was to have a well-run IRS [which] is critical to the operation of our tax system. From a public administration paradigm, the IRS moved from a traditional tax management methodology to a New Public Management (NPM) methodology. Traditional tax management focused on efficiency and effectiveness typically measured by tax collections per dollar spent. NPM focuses on performance based activities emphasizing the responsiveness to the needs of taxpayers. Ultimately, the IRSs core function remains the annual collection of nearly $2 trillion in taxes. The effects of the 1998 IRS Reform Act on the IRSs administration were assessed within the context of this paradigm shift in management. Partial replication of the IRSs 1990 Taxpayer Opinion Survey showed taxpayers perceptions of the IRS have not improved. In fact, the IRSs receipt of top quality service evaluations universally decreased. Analysis of selected enforcement/compliance data showed that the number of IRS auditors is inversely related to the annual tax gap. The data of this dissertation, however, indicated that the 1998 IRS Reform Act resulted in better taxpayer compliance. Examination of the IRSs personnel data shows that IRS executives have not received private pay parity, and the IRS strategically misrepresented §1203 employee termination violations in 2003.

In this study, I develop a measure of earnings quality by using qualitative characteristics of financial statement information specified in the Statement of Financial Accounting Concepts (SFAC) No. 2 (FASB 1980). I derive a summary measure of earnings quality by applying factor analysis on fifteen variables representing different components of two primary dimensions of earnings quality: relevance and reliability. I then test the validity of the earnings quality construct by examining whether the construct reflects decision usefulness that is operationalized in two ways: value relevance and cost of capital analyses. I provide empirical evidence suggesting that the earnings quality construct reflects decision usefulness to investors, which is consistent with the FASB's assertion. Finally, I explore the relative desirability of each dimension in light of decision usefulness of earnings information, and find that investors, in general, prefer the relevance to the reliability dimension of earnings.