Gerrymandering is a form of voting manipulation whereby electoral district boundaries are drawn to produce a partisan or political bias in elections. In this paper, we study partisan gerrymandering in the United States to understand its undemocratic outcomes and how the design of election institutions can promote or prevent gerrymandering. We begin with a survey of the history of gerrymandering, with a particular focus on partisan
gerrymandering. We then consider the normative standards of fairness in democracy that partisan gerrymandering may violate. Next, we present a typology of partisan gerrymandering based upon the district maps drawn in California, Pennsylvania, Massachusetts, and New Jersey for the 2012 elections. Using the partisan symmetry method, we estimate the seats/votes curves of the congressional maps used in 38 states during the 2012 elections. We find that partisan gerrymanders occur almost exclusively when politicians are given control over redistricting. This analysis implies that a political
designer, who wants to minimize gerrymandering, should not put redistricting in the hands of politicians.

The aim of this paper is to investigate the fi rms’ incentives to engage in process R&D under vertical industrial setting, when the raising rivals’ cost effect is present. We show that R&D investment of the downstream duopoly fi rm raises the rival’s marginal costs of production. The downstream R&D behavior can give rise to the symmetric investment games, i.e., the prisoner’s dilemma, the deadlock game and the harmony game, between downstream competitors. If the costs of the R&D investments made by the downstream fi rms are large enough, the downstream fi rms can participate in the harmony game, which results in the investment hold-up or the creation of the R&D-avoiding cartel. For more R&D-effi cient downstream fi rms, the downstream investment game can end up in the prisoner’s dilemma or the deadlock game. In the prisoner’s dilemma, both downstream fi rms invest in R&D, but such a behavior is not Pareto optimal. In the prisoner’s dilemma, greed and fear make fi rms invest in R&D. In the deadlock game, both downstream fi rms invest in R&D, and such a behavior is Pareto optimal. The R&D investments are not induced by any social tension (greed or fear).

Debate regarding evidence based medicine in current state focuses on basic philosophical as-sumptions of paradigm announced in 1992 by EBM Working Group. One of the topic in this debate is clinical reasoning in light of medical decision-making. Polish literature still lacks of comprehensive presentation in this theoretical fi eld. In current article I present overview of models of clinical reasoning. I begin from the debate of rationality distinguishing two positions called meliorism and panglosianism. Next I proceed to elaboration of three models for clinical reasoning in context of meliorists position. It gives an opportunity to look closer for inferential rules of clinical decision-making and as a result may be precious perspective for future re-search and medical education.

The article proposes a new type of statistical dependence. Statistical independence of this type means the identity of the frequency order of category choices in all groups. The compatibility between choices will be assessed by means of a diversity measure. A new measure of statistical dependence, its properties and interpretation will be specifi ed.