Science generates knowledge about the social and natural environment. This knowledge is always imperfect, because the instruments at hand to analyse the facts, be they social or natural, improve with the passing of time; moreover, things may be more complex than they seem. Nevertheless, the scientific knowledge available at any given moment is the best basis on which to make informed collective decisions. Ignoring this knowledge can be costly.

There is a clear consensus among scientists on many issues. This consensus is reached by conducting and replicating academic research studies. In these cases, an opinion that is contrary to scientific knowledge can be qualified as an unfounded belief.

Unfounded beliefs exist in both social and natural sciences. In the latter, unfounded beliefs include negation of climate change, the assertion that vaccines cause autism, and denial of the existence of the coronavirus. In the sphere of social sciences, the proposal to regulate rents by imposing an upper limit or ceiling is very popular as a measure for providing more people with access to housing (Brandts et al., 2021). However, it is an idea that lacks a scientific basis (Diamond et al., 2019).

Access to housing is a serious social problem today. Concern for social justice in access to housing may explain the popularity of the measure of rent capping in cities with tight housing markets. The proposal is attractive because it is simple, and would appear to enable many low-income families to gain access to housing. However, the great majority of scientific studies on the experience in cities in a variety of countries where rent controls have been implemented, show that the results have been disappointing. Not only has it failed to facilitate access to housing but has actually hindered it by creating new problems, such as long waiting times (up to 20 years in Stockholm), a rental black market and corruption. On the other hand, alternative housing policies exist that achieve the desired objective without creating these problems. One of them is to increase the stock of public rental housing.

The aim of this research is to ascertain whether the prevalence of this unfounded belief diminishes after scientific information is provided in two different ways. In order to evaluate the differential impact of these two ways of providing information, an experiment is conducted along the lines of clinical trials in medicine, with the participants divided randomly into two groups. The experiment consists in giving the participants in each group a different text with the same scientific information on the rent cap policy. The texts differ in the style adopted: one of the texts is refutational, while the other is expository.

The experiment is conducted online, and is organised as follows: at the beginning a questionnaire is distributed in order to elicit the opinion of each participant on issues related to housing. The participants then read the corresponding text explaining what rent capping means and what effects it has, according to research to date. Afterwards, participants answer a set of questions on the text and give their opinion again as to whether they believe this measure facilitates access to housing. The sample comprises 702 people living in Spain. Half of them are women, and the average age is 32; 44% of them have higher education and 19% live in rented accommodation.

Before reading the information about scientific studies on this issue, 78% of the participants agree with the statement that rent capping would enable more people to access housing. Only 13% disagree, while don’t-knows stand at 9%. Support is, therefore, mainstream (figure 1.1). There are only minor differences in the degree of agreement by gender, educational level, age and housing tenure status, as can be seen in figures 1.2 to 1.5.

In order to ascertain whether the prevalence of this belief diminishes after scientific information is provided in two different ways, each group of participants is given one of the two types of text to read. Using neutral language, the expository text develops the arguments and the evidence afforded by scientific studies on the negative effects of rent controls. In the refutational text, on the other hand, the misguided belief is explicitly challenged. Moreover, the language of the refutational text is intended to encourage reflection on the negative consequences of this measure. It recognises that concern for welfare and social justice in access to housing is an important reason for supporting rent controls, and it explains alternative policies that do work. The design of the refutational text is based on the recommendations of, among others, Tippett (2010), and has been used with some success to combat negation of climate change (Lewandowsky, 2021). Our main hypothesis, based on this previous literature, is that the refutational text would have a stronger effect than the expository text in reducing the belief.

The results allow us to observe that providing scientific information on the effects of rent controls lead to reconsideration of the unfounded belief and weaken it.

Figure 2.1 shows that, in the group who read the expository text, 78% of the participants initially agree with the belief. But after reading the text, only 41% agree with it. The belief therefore diminishes by 37 percentage points, i.e., almost half of those who held it initially.

Figure 2.2 shows that, in the group who read the refutational text, 79% of the participants initially agree with the belief. But after reading the text 41% agree with it. The belief diminishes by 38 percentage points, again almost half of those who held it initially.

The two texts – expository and refutational – therefore have a substantial and very similar effect. The regression analysis, which takes into account the different socio-demographic characteristics of the participants (gender, age, educational level, etc.), confirms that the effect of the refutational text is not significantly different from that of the expositive one.

The popularity of many unfounded beliefs is explained in part by the fact that they tend to be simple and intuitive, and adopting them reassures us in the face of a complex problem that disquiets and worries us. Various psychological factors contribute to the perpetuation of these beliefs, even when corrective information is received. One of them is the greater or lesser tendency towards reflective thinking, which requires a certain amount of effort, or towards instinctive thinking, which requires less effort. The latter predisposes one’s opinions or judgements on a subject to reflect cognitive biases (Kahneman, 2011).

Consequently, debunking unfounded beliefs is no simple matter (Brandts et al., 2021). There is a test to measure the propensity towards reflective as opposed to instinctive thinking, on a scale from 0 to 10. According to the score obtained in the test, participants On the basis of the distribution of the answers given, the participants are divided into four categories corresponding to the four quartiles of the distribution, from Q1 to Q4, as shown in the figures. Participants in quartile Q1 are the most instinctive, and those in quartile Q4 the most reflective.

Figure 3.1 shows how the initial opinion of the participants is distributed according to greater or lesser propensity towards reflective thinking. The degree of agreement or disagreement with the idea that a rent cap would encourage access to housing does not depend on having a greater or lesser reflective disposition.

However, changes in the belief do depend on this disposition. Figure 3.2 shows that the percentage of participants who reduce their agreement with the belief after reading the text is higher the greater their propensity towards reflective thinking.

The regression analysis, which takes into account the influence of factors such as employment status, housing tenure status, age, gender and educational level on the change of opinion, endorses these conclusions more clearly. Without going into technical details, the results of this analysis indicate that a stronger tendency to reflective thinking leads to greater disagreement with the belief, by a magnitude equivalent to 63% of the average change of opinion.

Below we analyse whether reconsideration of the belief after reading the information provided is related to the participant’s personal opinion as to whether governments should guarantee equal opportunities and to the degree of trust in social sciences.

Figure 4.1 shows the change in the belief according to the degree of agreement with the statement on equal opportunities. Participants who agree with this statement are those with the largest reduction in their support for the belief about rent controls.

Figure 4.2 presents the change in the belief according to the degree of agreement with the statement on the usefulness of the knowledge generated by the social sciences. Most of the participants who are most in agreement with this statement are those with the largest reduction in their support for the belief that setting rent caps increases access to housing. The figure also reveals that, albeit to a lesser extent, some of these participants increase their support. The regression analysis, more complete than a graphic description (as it takes into account the effect of other variables that may influence the belief about the effectiveness of capping), confirms that the greater the participants’ trust in the usefulness of the social sciences, the more they accept the scientific information they are given.

This research confirms the existence of an unfounded but very widespread belief that setting rent caps enables more people to access housing. The results show that providing scientific information helps to reduce the popularity of this belief. However, the communication formats we use are only partly successful, because 41% of the participants persist in the unfounded belief.

This study delivers evidence that giving the same information in two different ways – one expository and the other refutational – has very similar effects. Higher propensity towards reflective thinking helps to increase the impact of the scientific information. The reduction in the degree of agreement with the belief is greater among those who support the idea that governments should guarantee equal opportunities. Furthermore, the reduction is greater among those who have more trust that the social sciences are a good foundation on which to build laws and regulations.

Overall, this research highlights the importance of conveying the knowledge generated by the social sciences to the general public, in the same way as it is important in the case of the natural sciences. One aspect that we do not explore in this research is that the short-term effect of providing written information may be different from the long-term effect. In addition, other ways of communicating may have a greater impact. These are important issues which we propose to address in the future.

BRANDTS, J., I. BUSOM, C. LOPEZ-MAYAN and J. PANADÉS (2021): «Dispelling misconceptions in economics », Barcelona GSE Working Paper 1096.

DIAMOND, R., T. MCQUADE and F. QIAN (2019): «The effects of rent control expansion on tenants, landlords, and inequality: evidence from San Francisco», American Economic Review, 109(9).

KAHNEMAN, D. (2011): Thinking, fast and slow, Nova York: Farrar, Straus and Giroux (traducción al castellano: Pensar rápido, pensar despacio, Barcelona: Debate, 2012.

LEWANDOWSKY, S. (2021): «Climate change, disinformation, and how to combat it», Annual Review of Public Health, 42.

TIPPETT, D. (2010): «Refutation text in science education: a review of two decades of research», International Journal of Science and Mathematics Education, 8(6).