Second-order Bayesian revision of a generalised linear model

It is well known that the exponential dispersion family (EDF) of univariate distributions is closed under Bayesian revision in the presence of natural conjugate priors. However, this is not the case for the general multivariate EDF. This paper derives a second-order approximation to the posterior likelihood of a naturally conjugated generalised linear model (GLM), i.e., multivariate EDF subject to a link function (Section 5.5). It is not the same as a normal approximation. It does, however, lead to second-order Bayes estimators of parameters of the posterior. The family of second-order approximations is found to be closed under Bayesian revision. This generates a recursion for repeated Bayesian revision of the GLM with the acquisition of additional data. The recursion simplifies greatly for a canonical link. The resulting structure is easily extended to a filter for estimation of the parameters of a dynamic generalised linear model (DGLM) (Section 6.2). The Kalman filter emerges as a special case. A second type of link function, related to the canonical link, and with similar properties, is identified. This is called here the companion canonical link. For a given GLM with canonical link, the companion to that link generates a companion GLM (Section 4). The recursive form of the Bayesian revision of this GLM is also obtained (Section 5.5.3). There is a perfect parallel between the development of the GLM recursion and its companion. A dictionary for translation between the two is given so that one is readily derived from the other (Table 5.1). The companion canonical link also generates a companion DGLM. A filter for this is obtained (Section 6.3). Section 1.2 provides an indication of how the theory developed here might be applied to loss reserving. A sequel paper, providing numerical illustrations of this, is planned.     

Lean and mean in the civil service: the case of processing in HMRC

The public sector has been importing private sector methods and practices aimed at generating efficiencies and cost savings. However, the consequences of these changes on the working lives of civil servants are under-researched. This article uses detailed fieldwork to investigate the impact of Lean on labour processes in HM Revenue and Customs (HMRC). We argue that Lean has a detrimental effect on employees, their working lives, and the service that is provided to the public. The consequences of Lean on public sector work are highly problematic, which is of serious concern given its progressive impact on other civil service departments in the UK.     

Does the economy matter? Tough times,good times,and public service motivation

This paper describes a pilot study which explored the effects of different economic conditions on the public service motivation (PSM) levels of a small group of government and private sector employees in Malaysia. It address three questions: are the PSM levels of government employees significantly higher than those of private sector employees; are the PSM levels of employees significantly influenced by their perceptions of the external economic environment; and is the relationship between the economy and PSM affected by the sector of employment? Government employees’ PSM levels were found to be higher than those of the private sector employees, and were sensitive to perceptions of different economic conditions.     

FX intervention in the Yen-US dollar market: a coordination channel perspective

The coordination channel has recently been established as an additional means by which foreign exchange market intervention may be effective. It is conjectured that strong and persistent misalignments of the exchange rate are caused by a coordination failure among fundamentals-based traders. In such situations official intervention may act as a coordinating signal, encouraging traders to engage in stabilizing speculation. We apply the framework developed in Reitz and Taylor (Eur Econ Rev 52(1), 55?C76 2008) to daily data on the yen-US dollar exchange rate and on Federal Reserve and Japanese Ministry of Finance intervention operations. The results provide further support for the coordination channel of intervention effectiveness.     

Household Finances and Social Interaction: Bayesian Analysis of Household Panel Data

We investigate the relationship between social interaction and household finances using the British Household Panel Survey. We explore the relationship between a wide range of aspects of household finances and social interaction, rather than focusing on one particular facet of household finances, such as the holding of stocks and shares. We develop a Bayesian statistical framework to simultaneously explore both sides of the household balance sheet—liabilities and assets. Additionally, we allow the influence of social interaction on household finances to be time dependent, enabling us to model the effects of social interaction from a dynamic perspective. We also develop a two‐part model to jointly investigate the influence of social interaction on the amount of different types of debt and financial assets held conditional on holding the different types of debt and assets. Our analysis suggests that social interaction is associated with households holding larger amounts of debt and assets.     

How does qualitative data collection modality affect disclosure of sensitive information and participant experience? Findings from a quasi-experimental study

Quality & Quantity - Focus groups (FGs) and individual interviews (IDIs) can be conducted in-person or in several different online contexts. We conducted a quasi-experimental study and assessed...     

Testing for unit roots in the presence of uncertainty over both the trend and initial condition

In this paper we provide a joint treatment of two major problems that surround testing for a unit root in practice: uncertainty as to whether or not a linear deterministic trend is present in the data, and uncertainty as to whether the initial condition of the process is (asymptotically) negligible or not. We suggest decision rules based on the union of rejections of four standard unit root tests (OLS and quasi-differenced demeaned and detrended ADF unit root tests), along with information regarding the magnitude of the trend and initial condition, to allow simultaneously for both trend and initial condition uncertainty.     

Explaining Individual and Task Differences in Consumer Attitudes toward Ambiguity

Consumers frequently face choices involving ambiguity, or uncertainty about the probabilities of the various possible outcomes. Past research has found that attitudes towards ambiguity vary across tasks. It is the purpose of this paper to suggest that such variation in consumer attitudes can be explained by examining the nature of the source of the uncertainty. It is suggested that consumers will have different attitudes depending upon whether the ambiguity is generated internally (by the consumer) or externally (by some outside source). Exploratory evidence and propositions for future research are presented. The results suggest that marketing strategies and de-biasing methods for ambiguous situations should consider the source of the ambiguity.     

The Cash Recovery Method of Calculating Profitability: An Application to Pharmaceutical Firms

The problems with commonly used accounting profit rates are well documented. In this paper an alternative to accounting profit rates, the cash recovery method is investigated and improved. This improved method is used as a means to estimate profitability in the pharmaceutical industry on a firm level. The profitability estimates give a similar rank order to the accounting profitability rates, but have different magnitudes.     

Load and no-load mutual fund dynamics during the 1987 market crash: A stochastic dominance analysis

Using a large sample of equity mutual fund returns, we compare performance of load and no-load funds during the 1987 crash. Differences in return distributions, particularly in the higher moments when the market was under stress, suggest a greater use of portfolio insurance by no-load fund managers. Using stochastic dominance, we find that load and no-load funds performed equally well before the crash. No-load returns dominated load fund returns during the crash. Load fund returns dominate after the crash. Over the entire month, no-load funds dominate. We attribute this to investor behavior motivated by the lack of a front-end load.