Historical evidence for anchoring bias: The 1875 cadastral survey in Istanbul

In experiments where participants value objects, it is well known that their appraisals can be systematically affected by irrelevant information; this is known as the anchoring effect. We ask whether the anchoring effect can be found in a real-world event that took place in 1875, involving real property appraisals for the first cadastral survey in Istanbul. After controlling for all the available features, such as rental price, size, region, number of rooms, construction material, type of building, etc., we find a statistically significant positive relationship between the property appraisals and the door numbers of these properties. That is to say, two neighboring rented houses with identical physical features and the same rental price would be expected to be differently valued by the appraisers due to the difference in their door numbers. The house with a higher door number would have a higher appraised value. Because property appraisals were used to calculate property taxes in 1875, our result suggests that cognitive bias can have large-scale legal, and even fiscal, implications for the state.