Table 1. Demand for redistribution, regression analysis.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Cognitive ability	−6.665***	−5.026***		−4.264**		−4.238**		−4.819***		−4.415**		−4.846***		−3.696**		−3.475*
	(1.621)	(1.723)		(1.769)		(1.759)		(1.683)		(1.778)		(1.749)		(1.796)		(1.858)
Mean annual income			−0.235**	−0.178**									−0.138	−0.138	−0.139	−0.101
			(0.092)	(0.089)									(0.087)	(0.087)	(0.099)	(0.095)
Beliefs: luck/effort					−2.874***	−2.577***							−2.403**	−2.403**	−2.247**	−2.146**
					(0.990)	(0.985)							(0.992)	(0.992)	(1.051)	(1.048)
Beliefs: gov efficiency							2.188***	2.176***							2.231***	2.183***
							(0.747)	(0.751)							(0.750)	(0.751)
Risk aversion									0.876	0.911					0.661	0.736
									(0.896)	(0.877)					(0.877)	(0.864)
Altruism											−0.151	0.111			−0.097	−0.005
											(0.723)	(0.723)			(0.676)	(0.680)
Standard controls	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	271	266	266	266	265	265	262	262	247	247	262	262	265	265	243	243

OLS. Robust standard errors in parentheses. Significance levels: ***p<0.01, **p<0.05, *p<0.1. Dependent variable is demand for redistribution measured in percent where 0% =  no redistribution, 100% = full redistribution. Cognitive ability measure from military enlistment data with zero mean and unit standard deviation in the population of all enlisters. Standard controls are age, education level (dummy for primary school being highest education level), and self-reported socio-economic status during childhood (alternatives "working class", "lower middle class", "middle class", "upper middle class" and "upper class"). Mean annual income is measured in SEK 10,000 s, mean taken over the years 1999 to 2010. Beliefs about luck/effort measured with question in survey about what matters (luck or effort) for how well an individual does economically in life, higher number indicates more importance for effort. Beliefs about government efficiency measured in survey with question about how efficiently the public sector in Sweden redistributes resources so that no resources get lost on the way, higher number indicate more efficiency. Risk aversion measured in survey with eight hypothetical choices between a fixed amount and a lottery, risk aversion is sum of choices of the fixed amount. Altruism assessed in survey with hypothetical question about the willingness to give to charitable purposes (1 = willing to give, 0 = not willing to give). All results are robust to using Ordered Probit (OP) instead of OLS (results available in Table S1 in SI).