The bias of least squares polynomial interpolants

The fitting of a straight line — or more generally of a low degree polynomial — to a point cloud in the plane is a commonly performed statistical technique. This paper discusses the bias of such a procedure and in particular generalizes the well-known remainder formula for polynomial interpolation to the regression setting. Designs minimizing the maximum bias are discussed as well.     

Face to face and telephone surveys in terms of sampling representativeness: a multidimensional analysis

This paper analyzes sample structure in a survey-based research project using both face-to-face and telephone survey techniques, at a time when the telephone survey has completely ousted the face-to-face survey; particularly in electoral studies. Multiple Correspondence Analysis is applied to the data from a political poll conducted prior to regional elections held in Spain’s Autonomous Community of Galicia in which half of the questionnaires were administered by telephone and the other half via face-to-face interview. Unlike other studies—that use a partial approach based on bivariate analysis—this one examines all the sociodemographic variables simultaneously.     

International patterns of cigarette smoking and global antismoking policies

This paper provides a recent look at global smoking trends across population subgroups for a large sample of countries. Attention to international aspects has been scarce in the literature. The prevalence of various smoking control policies across different country groups is discussed. The level of economic development in a country and the demographic characteristics of the population appear to be crucial determinants of smoking prevalence. Among four categories of smoking control policies—advertising restrictions, sales restrictions, territorial restrictions, and packaging requirements—territorial restrictions on tobacco use seem most popular across countries, while sales restrictions are least popular. Price (tax)-based policies to control smoking are not effectively being used in many countries, especially developing countries.     

Mail survey abroad with an alternative web survey

The large decline in cooperation in surveys experienced in the first decade of the twenty first century has placed representative surveys in a veritable quagmire, although numerous researchers have shown that the decline in the total nonresponse has been much higher in administered surveys (face-to-face and telephone) than in self-administered (mail and web surveys). The move from the “traditional” mail survey to the internet survey has brought important changes, but, at the same time, new challenges have arisen related to the lower coverage and the reduced visibility, the need for certain skills to fill out a questionnaire, changes in response rates, etc. This is a self-administered survey that informs the addressees by mail that they have been selected to participate in a research survey to which they can respond either by using the paper questionnaire (in the letter there is a prepaid envelope), by means of a web questionnaire (accessing a link and using a password), or by telephone (the respondent provided a contact time telephone number via a prepaid envelope or by calling an answering machine). The respondents themselves, therefore, choose the mode of response. An analysis of the responses achieved by each modality shows that the majority of the questionnaires were received by ordinary mail (83.6 %), 359 by internet (14.4 %), and 51 were answered by telephone (2.0 %).     

A New Indicator of European Integration and an Application to Collaboration in Scientific Research

The goal of this study is to develop a comprehensive indicator of integration among countries within a supranational system. Integration is not analysed in terms of the growth in interactions among countries, but in terms of the matrix distribution of interactions among countries. Integration can then be indicated in terms of interaction biases among countries measured by the difference between the observed matrix distribution and the hypothetical random distribution. The indicator is applied to data on research collaborations among European research institutions (1993-2000). Evidence is found that the European science system has indeed become more integrated. The higher level of integration has resulted exclusively from a more evenly distributed pattern of European collaborations, while the strong bias towards intra-national collaborations persisted. The results point to the persistence of national science systems. A future research agenda and science policy implications are discussed.     

FDI forward spillover effects in emerging markets: A comparative meta-analysis of China and Eastern Europe

Using two unique datasets of 530 forward estimates from 19 studies on China and 942 forward estimates from 28 studies on Eastern European transition economies, this study quantitatively analyzes foreign direct investment forward spillover effects, accounting for various biases and firm attributes in existing empirical results using a meta-regression analysis. We find that aggregation bias, misspecification bias, endogeneity bias (including simultaneity bias and omitted-variable bias), and publication bias, as well as firm attributes, contribute to the heterogeneity in forward spillover estimates; and that the preferred forward spillover effects are positive but nonsignificant. Interestingly, the preferred forward spillover effect is likely to be much larger in Eastern European transition economies than in China.     

Equity aspects of VAT in emerging European countries: A case study of Serbia

A belief that consumption taxation is inherently inequitable has been entrenched in a significant portion of the general public and was supported by early empirical evidence that suggested a highly regressive annual VAT incidence. However, it has been shown that much of the estimated annual VAT regressivity is due to the income under-reporting bias inherent in sample surveys. This bias is particularly important in emerging European countries due to a high shadow economy and the evasion of direct income taxes, which suggests household expenditures as a more meaningful indicator of well-being than registered income. Furthermore, theoretical considerations favor the lifetime incidence approach, whereby VAT is estimated to be proportional or mildly progressive. A micro-simulation analysis of the Serbian expenditure survey data yields incidence estimates in line with the existing literature from other countries. We show that a significant presence of own-source (small) farming production of food in many emerging European countries, including Serbia, presents an important progressivity-enhancing buffer compared to the VAT incidence in developed European countries. We conclude that the common beliefs of inherently inequitable VAT taxation are vastly overstated and poorly founded in the economic reality of emerging European countries such as Serbia, where VAT can be most adequately described as being mildly progressive.     

The effect of environmental innovations on employment changes: an econometric analysis

This paper examines the determinants of employment changes due to an environmental innovation of an establishment. The data stem from telephone surveys in five European countries. 1594 interviews have been realized with environmentally innovative establishments representing the European industry and service sector. Based on results of discrete choice models, we show that if the most important environmental innovation is a product or service innovation it has a significantly positive effect on the probability of an increase in employment compared with the probability of no noticeable change. In contrast, if the most important environmental innovation is an end‐of‐pipe innovation it has a significantly positive influence on employment decrease. Methodologically, we consider the multinomial logit model and several multinomial probit models. We find that the estimates of the parameters of the explanatory variables are very similar in the different approaches. Copyright © 2004 John Wiley & Sons, Ltd and ERP Environment.     

The dynamics and distribution of the area price in the Nord Pool

The Nord Pool is often cited as a standard and successful electric power exchange. It was first created in Norway and developed into the power change system covering the Nordic countries. The Nord Pool provides a physical market where electricity producers and consumers/distributors meet by submitting bids for sale and purchase for 24 h time segment. If there does not exist a restraint of transmission capacity across the member countries, a single price—system price—is calculated by the intersection of the demand and supply curves, whereas if there exists congestion of transmission lines, area price instead of system price emerges. We are interested in analyzing how often and how far the two prices diverge and the probability distribution of the price ratio between system and area price. We found that the price ratio does not follow the normal distribution but the distribution is fat-tailed.     

The potential of a multi-mode data collection design to reduce non response bias. The case of a survey of employers

The aim of the paper is to compare two alternative survey designs in terms of resultant response rates, non response bias and cost. The first design is a simple postal survey with follow-up mailings; the second design is a two-phase multi-mode design, where the postal survey is followed at the second phase by a telephone survey of non-respondents. We present a case study based on a survey of employers. In this study we find evidence that the sample obtained using only postal methods is biased in important respects. Bias is not apparent in the demographic characteristics of the employees. But bias is observed in some of the employees’ employment characteristics and some of the characteristics of the firms in which they work. The multi mode design seems, overall, to have reduced or removed the bias of the postal sample. Only in marginal respects was some further bias introduced. We also compare costs of the two designs, to enable a comparison of cost-effectiveness at bias reduction.     

DOES ICT GENERATE ECONOMIC GROWTH? A META‐REGRESSION ANALYSIS

Despite phenomenal technological progress and exponential growth in computing power, economic growth remains comparative sluggish. In this paper, we investigate two core issues: (1) is there really no connection between ICT and national economic growth? and (2) what factors moderate the ICT–growth relationship? We apply meta‐regression analysis to 466 estimates drawn from 59 econometric studies that explore the Solow or Productivity Paradox that there is little impact of ICT on economic growth and productivity. We explore the differential impact of ICT on developed and developing countries and the differential impact of different types of ICT: landlines, cell phones, computer technology and Internet access. After accommodating potential econometric misspecification bias and publication selection bias, we detect evidence that ICT has indeed contributed positively to economic growth, at least on average. Both developed and developing countries benefit from landline and cell technologies, with cell technologies’ growth effect approximately twice as strong as landlines. However, developed countries gain significantly more from computing than do developing countries. In contrast, we find little evidence that the Internet has had a positive impact on growth.     

The Dirty Dozen: Common Problem Areas in Principal–Agent Relationships

The main purpose of the article is to educate managers about potential difficulties caused by breach of fiduciary duties on the part of agents/employees within their organization. Managers will learn how to recognize common problem situations so they can take steps to avoid, or at least mitigate, any resulting damage. This article highlights 12 frequent situations—the “dirty dozen”—that pose dangers for business organizations.     

Examining Turkey and member states of European Union in terms of health perspectives of millennium development goals

Development is related not only to production and increase in per capita income but also to social, cultural and political improvements. The purpose of development is that individuals would live long, healthy and happy lives thanks to economic development of society. From this perspective, it is obvious that the human factor is fundamental to the concept of development. This study examines the most important element in human development—health. As health indicators, it uses the health perspectives in the United Nations millennium development goals that are “reduce child mortality”, “improve maternal health” and “combat HIV/AIDS, malaria and other diseases”. In the study, European Union (EU) member states and candidate country Turkey are compared in terms of health related goals among millennium development goals by using Multidimensional scaling and Cluster analysis. Initially, countries with similar and dissimilar health indicators are mapped in multidimensional space by multidimensional scaling analysis. Further, the same method is used to reveal similar and dissimilar health indicators among countries. Then, the findings are compared using Cluster analysis and are identified to be similar.     

Institutional barriers in labor markets: Examples, impacts, and policies

This paper examines the institutional biases that impede the competitive functioning of labor markets. Two contexts are considered. The first relates to Moroccan labor migrants in The Netherlands, where institutional bias distorts the competitive functioning of the labor market by downgrading the educational returns to migrant workers and acting as a disincentive for further investment in human capital. The second relates to labor markets in Indonesia and Pakistan. Institutional bias in these two countries leads to an exaggeration of labor returns to certified education, and to over-investment in university education. We argue that such biases are fed by misinformed beliefs and group interests, and stand in the way of achieving higher growth and equity.     

Trade intensity,country size and corruption

Abstract. Several authors claim to provide evidence that governmental corruption is less severe in countries where trade intensity is higher or populations are smaller. We argue that theory is highly ambiguous on these questions, and demonstrate that empirical links between corruption and trade intensity – or country size, strongly related to trade intensity – are sensitive to sample selection bias. Most available corruption indicators provide ratings only for those countries in which multinational investors have the greatest interest: these tend to include almost all large nations, but among small nations only those that are well-governed. We find that the relationship between corruption and trade intensity disappears, using newer corruption indicators with substantially increased country coverage. Similarly, the relationship between corruption and country size weakens or disappears using samples less subject to selection bias. Received: July 2001 / accepted: April 2002 We thank Anand Swamy and two anonymous referees for helpful comments and suggestions, Paul Schorosch for able research assistance, and Ray Fisman, Roberta Gatti, Aart Kraay, and Shang-jin Wei for kindly providing data. The conclusions of this paper are not intended to represent the views of the World Bank, its Executive Directors, or the countries they represent.     

The management of an ageing workforce: organisational policies in Germany and Britain

Demographic change as well as pressure from the European Union and national government are forcing organisations to change age‐discriminatory HRM approaches. Based on a qualitative analysis of eight British and German organisations, we found that commitment, scope, coverage and implementation of age management differ due to country‐specific institutions, particularly government, in nudging employers and unions to preferred age practices. This confirms the path dependency concept suggested by institutional theory. Nevertheless, we also found that industry‐specific factors mediate the implementation of age management, leading to some convergence across countries. This indicates that organisations deviate from the institutional path to implement practices that they deem important.     

Subperfect Game: Profitable Biases of NBA Referees

This paper empirically investigates three hypotheses regarding biases of National Basketball Association referees. Identification of basketball referee bias is typically difficult as changes in observed statistics may be caused by either changes in referee bias or player behavior. We identify bias by exploiting the fact that referees have varying degrees of discretion over different types of a particular statistic‐turnovers. This allows us to conduct a treatment and control‐style analysis, using the less discretionary turnovers as the player behavior control. The results provide evidence that referees favor home teams, teams losing during games, and teams losing in playoff series. All three biases are likely to increase consumer demand.     

Global cultural zones the empirical way: value structure of cultural zones and their relationship with democracy and the communist past

The paper empirically identifies the world’s cultural zones, comprised of the countries with the distinctive patterns of values, and determines their relationship with the country’s communist past and current state of democracy. The analysis is based on the countries’ average values on the two main dimensions of cross-cultural variations: the traditional versus secular-rational and the survival versus self-expression values. The World Values Survey and European Values Study data from 101 countries were used. The K-means method of cluster analysis and the Variance Ration Criterion for determining the final number of clusters were performed. The solution with three clusters appeared as the most appropriate and theoretically meaningful, representing three broad cultural value syndromes. The first value syndrome was defined by low self-expression and high secular-rational values (the zone of secularized authority), the second one by low secular-rational and low self-expression values (the zone of traditional authority), while the third was characterized by high values on both value dimensions (the zone of emancipation). The countries’ cluster membership is highly correlated with the (non-)communist past and the level of socioeconomic development. The countries in three cultural zones significantly differ in terms of the state of democracy: democracy is more fully developed in the countries in the cultural zone of emancipation, and the least developed in the countries with the prevailing traditional authority cultural outlook.

     

Face-to-face versus telephone surveys on political attitudes: a comparative analysis

Several researchers declared that the telephone survey reaches more accurate voting tendencies than the face-to-face surveys. Telephone survey shows numerous advantages compared to a face-to-face one but, however, the telephone survey also has some inconveniences. Among these it is important to highlight the scant quality of the sampling frame; absence of a telephone in some homes and the wide expansion of the mobile phone; low response rate of certain collectives and the overrepresentation of others. There are also some new barrierswhich make access more difficult (e.g. the automatic answering phone) and the saturation of the telephone medium because of the large amount of publicity activities which generate a large number of “unsuccessful” calls and interruptedinterviews. The objective of this paper is evaluating the adaptation of the telephone surveys in the electoral forecasts; in an attempt to see if it shows substantial improvements when they are compared to face-to-face interviews.     

An examination of estimation and specification error biases in estimates of option prices generated by Black-Scholes and Cox-Ross models

In this paper the author identifies and examines the estimation and specification error biases of the Black-Scholes and Cox-Ross models by using both analytical and monte-carlo simulation techniques. Several hypotheses are tested. The central hypothesis is whether or not the estimation error bias in the correctly specified model is large enough to make researchers mistakenly pick the “wrong” model as being the “correct” one. The findings in this study support this hypothesis. It is shown that there is a bias toward accepting the Cox-Ross model as correct, even if the Black-Scholes is assumed to be the true model for pricing options.