Replication Data for: Comparing Three Measures of Legislative Professionalism

We evaluate three measures of state legislative professionalism: Squire’s (1992) index that measures professionalism relative to the US House, Bowen and Greene’s (2014b) two dimensional scaling, and legislative operating expenditures per member, an older measure that remains in occasional use. Replications of 18 recent articles show that these three measures regularly produce significantly different estimates of the effect of professionalism, particularly in longitudinal analysis; when they do, the choice of measure often affects whether other central variables retain a significant relationship with the dependent variable. These divergent results appear to reflect differences among these measures in terms of missingness, vulnerability to outliers, measurement of session length, and whether to benchmark to the US House. Researchers seeking a general indicator of professionalism should consider these differences when choosing an appropriate measure.本研究针对三类州议会职业化程度测度指标展开评估：一是斯夸尔（Squire, 1992）指数，该指标以美国众议院为参照基准衡量职业化程度；二是鲍恩与格林（Bowen and Greene, 2014b）提出的二维量表法；三是人均议会运营支出，这是一种仍偶有应用的传统测度方式。对18篇近期学术文献的重复验证结果显示，上述三类测度指标往往会得到存在显著差异的职业化程度影响效应估计结果，在纵向分析中这一差异尤为突出；当出现此类分歧时，测度指标的选择通常会直接影响其他核心变量与因变量之间是否仍能保持显著关联。此类结果分歧本质上源于三类测度指标在数据缺失率、异常值敏感性、会期时长测度逻辑，以及是否以美国众议院为参照基准等维度上存在差异。若研究者旨在获取通用性较强的职业化程度测度指标，在遴选适配的测度方式时应充分考量上述差异。