Vote Intention, Social Desirability Bias, and AfD: How Alternative Measurement Techniques can Improve Political Research

In Germany, the standard vote intention survey item has come under attack because it failed to correctly measure the vote share of the German party "Alternative fur Deutschland" (AfD). We argue that alternative measurement techniques that aim to reduce social desirability bias are better suited for this task. We test three measurement techniques to forecast AfD vote share-a double list experiment, the crosswise-model randomized response technique, and the wisdom of crowds design-and compare their performance to the standard vote intention item. Our results indicate that the wisdom of crowds design is an easily implementable and promising addition to political scientists' toolbox of survey items.