Simon Hug, "Studying the Electoral Success of New Political Parties: A Methodological Note," Party Politics, 6 (April 2000), 187-197.

First Paragraph:
New political parties of different types have attracted increasing attention over the last few years. Scholars have studied in detail green or left-libertarian parties (e.g. Kitschelt, 1988; Muller-Rommel, 1993), new right-wing parties (e.g. Ignazi, 1994; Kitschelt, 1995; Muller-Rommel, 1998), or regional and non-state-wide parties (De Winter, 1995; De Winter and Tursan, 1998). An important aspect of most comparative studies is often the explanation for the electoral success of newly formed parties. If scholars employ countries as units of analysis, they classify them according to whether a particular type of party has been electorally successful or not (e.g. Harmel and Robertson, 1985; Kitschelt, 1988, 1995; Kalyvas, 1996). If the new parties appear as units of analysis, scholars attempt to explain variation in electoral success with a wide array of factors (e.g. Muller-Rommel, 1993, 1998).

Figures and Tables:
None.

Last Paragraph:
While this example uses a particular research strategy, namely focusing on new parties that have emerged, other approaches also have to deal with this particular variant of selection bias. Whether a researcher compares countries with or without successful new parties, employs case studies or focuses on parties that have emerged, problems of selection bias are likely to appear. Instead of neglecting such problems, this note suggests that it is often beneficial to directly address the selection process that leads to the formation of new parties. Often interesting and important substantive results appear in such analyses, which would otherwise remain undetected.