Social policy - papers

Christensen et al., 2024

Rich people don’t like poor people near them

__________

A field experiment (FE) study examining whether rising local inequality increases support for redistrubution among high income groups.

It doesn’t: Increased inequality in ones local areas, reduces support for redistribution among the richest quantile.

Field experiment, DK

#Inequality
_________

RQ

Does rich people become more or less supportive of redistribution when exposed to poor people

Christensen et al (2024) sets off to determine whether rich people become more supportive of redistribution (as argued by the contact explanation) or less supportive (as argued by Sands in the conflict explanation).

Expectations

Contact explanation

-        Exposure to outgroup à sympathy à more support for redistribution

Conflict mechanism

-        Conflict perspective – exposed to outgroup à more negative view of that group.

-        Ritch person sees poor person

o   Insecurity – I could be like this, I need to keep my money

o   Derservingness – they do not deserve redistribution.

Christensen et al critised these mechanism, saying they either can be caused by self-selection or single or repeated exposure to poor people.

Contact explanation: Self-selection

-        Most contact studies are based on observational data – high risk of selection.

-        Rich people are cable of deciding where to live. Thus the results could partly be explained by rich people more positively disposed to poor people are more likely to live in areas with a higher fraction of poor people, than those rich folks who don’t like poor folks.

Conflict explanation - Single vs. repeated exposure

-        The Sands experiment is only ONE exposure to a poor looking guy in a righ neighborhood. The difference might be explained by different effects of a single exposure and a long-term exposure

Method

Panel study some fine-grained location data.

X: Changes in local income-composition.

Within model

-        Twoway fixed effects

-        Control for time invariant confounding

-        Looks at the difference within the individual over time.

Between-individual model

-        Just normal cross section analysis I think

-        Differences between individuals.

Results

Within-individual model

-        Exposure to poor people à less supportive of redistribution

-        Only for the richest quantile.

Between-individual model

-        Exposure to poor people à more supportive of redistribution

The within-individual model is best, thus indication support for the self-selection theory - that those rich people who likes redistribution lives in areas with higher income inequality. but if you look at a right individual who over time experience increased income inequality, they become less supportive of redistribution.

Thoughts

Interesting as it test a quite famous experiment (the sands experiment) in a more real-life setting, looking at real income changes in ones neighboorhood.

Mullin and Hansen, 2023

Fear is good, fear gets big things done

Survey experiment shows that if local politicians are informed that local new writes about shitty infrastructure, they are more likely to support infrastructure projects, because they won’t fear public punishment for using a lot of money as much. Surveyexperiment, US #Infra structure

RQ

What can motivate a politician to support infrastructure investment? Because they won’t necessarily do this cause low incentives.

Infrastructure projects = High-cost short term, high-benefit long term —> Politicans fear electoral punishment for using a lot of money short term.

The opportunity to be rewarded for problem-solving balances against the electoral punishment the politician expects for the investment’s short-term costs. Publicly available information can therefore reduce the constraint on a politician’s decision making.

Expectations

Higher support for investment if infrastructure failure is covered by the local news.

Higher support for investment if competition from other politicians

Higher support for investment if ambitious.

Method

Survey experiment

Treatment:

• Frame about a expensive project to replace a water pipe

• TG = Problem is covered by the local press

• CG = Problem is not covered by the local press

• Y = support for the project

N = 657

Unbalanced on some covariates – more democrats than rep

Results

We found that news coverage increased support for investment, but only for politicians who had faced an opponent in their previous election.

Not support for ambition

Thoughts

Hilbig and Wiedemann, 2024

*Housing construction reduce local government support because it just ain’t that important*

Municipality level Electoral punishment when building public housing because people generally prefer using the public money on something else – especially in municipalities where housing is not expensive. Dif’n’dif + survey, Germany #housing construction/NIMBYism

RQ

How does new public housing construction (X) affect incumbent vote share in municipalities (Y) in Germany.

Motivation

Tight housing market.
Problems with getting politicians and people to support new housing.

But does building new public housing actually decrease support for local governments?

Public housing is especially tricky as they per design attracts lower income groups, typically with some issues etc.

Expectations

New public housing construction generally decreases incumbent vote share

Testing 3 mechanisms

-        Homeowners oppos public housing because of worries of it causing the property values to decline by changing neighborhood characteristics etc.

-        Native local residents (not only homeowners) vote against public housing because they perceive they inhabitants (low income, immigrants of these homes as undeserving.

-        People are against construction of new public housing because they prefer spending the public money on something else.

Method

Dif’n’dif and some survey

Treated: municipality where public housing is build

Untreated: municipality where public housing is not yet build (but decided)

Results

Finds negative significant effect of public housing building and incumbent vote share across different measures (per capita, log etc.). However, the authors note that they are small.

Larger electoral punishment when closer to elections and something.

The mechanism finding support is that people prefer spending public money on something else than public housing. This is especially the case for municipalities with cheap housing, whereas municipalities with more expensive housing are more willing to prioritize construction of public housing (because they probably need it).

Thoughts

Relation to my synopsis

-        Hmm, not really. Just that housing policy, generally, influence party preferences.

Kreitzer et al., 2023

50 shades of deservingness: Texans likes gunowners better

Deservingness and its (sub)state level variations – deservingness perceptions of different groups (criminal, gunowners etc.) differs and also differs across states. Deservingness perceptions correlations with policy outcome. Observation, US #Deservingness

Kreitzer et al., 2023

Fifty shades of deservingness: an analysis of state-level variation and effect of social constructions on policy outcomes

RQ

How does deservingness differ across subnational levels and how does it relate to policy outcome.

Differs across the target group.

Expectations

Deservingness heuristics, despite of being a universal mechanism, differ across national and subnational levels because of

-        Policy feedback/path dependency: When given a social benefit to a target group, people tend to increasingly see that target group as deserving for that and more benefits. And vice versa for non-deserving

-        Elite and media deciptions of groups

-        State culture: different kinds of people live specific places kinda. “variation in state-level personality across the American states corresponds quite strongly with states core political characteristics”

H1: Subnational deservingness evaluations of target groups will vary across states.

H2: Subnational evaluations of deservingness will predict state-level policy outcomes.

Method

MTurk observational survey with poststratification

Y: “some groups are viewed less deserving of pity, sympathy or help than others. What do you think?” e.g. prisoners, gun owners, drug-users etc.

Results

Differs across states

To some extent predict state-level policy outcomes (but this is not causal, I would say).

Thoughts

The research of H2 is not causual but correlations.

Perhaps replicate in European countries?

Differences in stereotypes?? How sure are we that it is differences in deservingness and not different stereotypes that then trigger different deservingness? If that makes sense. Like in DK it “criminal” would not have the same racial connotation – perhaps leading to some racism to people with Arabic ethnicity.

Marantz & Lewis, 2022

Multifamily housing construction Size matters – concave (unhappy smile) relations between jurisdiction size and construction of multifamily housing. Observation, US #NIMBYism

RQ

Does size matter? In relation to construction of housing (multifamily housing)

Motivation

Tight housing market à need and demand of more housing.

However, the construction of multifamily housing is often meet by NIMBYism.

Hard for politicians to build new housing then.

Might depend on jurisdiction size – different reasons to why it should.

Expectations

The bigger jurisdiction à the more development of houses

Because (doesn’t test it)

-        More noticeable in small jurisdictions

-        Home owners are less likely to be local politicians primary constituency in big jurisdictions than in small suburban areas. Big jurisdictions have more and different interest groups

-        More renters in big jurisdictions

-        Politicians are in big jurisdictions more removed from local disputes à less responsive to NIMBYism

-        Real estate firms and construction unios have a bigger influence on politicians in big jurisdictions as campaigns are more expensive

Method

Observational

Y: multifamiliy housing constructed in areas

X: number of people in tract

Tract level data from US.

Like from 2010-2014 ish.

Results

The relationship between jurisdictional size and housing construction is non-linear (downward concave = ^ = unhappy smiley).

Meaning – in the smallest and biggest jurisdictions (< 1 mio), the less houses are built.

Thoughts

Vinæs and Kettle is a critique of this.

Contrary to Findor et al., Vinæs and Kettel argues that smaller jurisdictions may are more likely to build public housing in a way to increase the population size and gain a more tax revenue

Findor et al., 2023

The most deserving out-group goes to…. the reciprocity!

Deservingness – the Slovak-Roma experiment. In-group are more supportive of out-group benefits if reciprocity than if need or equality. The individual support for redistribution to out-group is higher than the perceived majority support Survey and field experiment, Slovania #Deservingness

RQ

What are effective ways to increase support for out-group redistribution?

Expectations

Testing if appealing to the three principles of distributive justice increase support for put-group redistribution:

-        Equality: everyone gets the same with no regard to reciprocity or need

-        Reciprocity: Social services conditional on effort.

-        Need: social services are targeted individuals with the largest marginal benefit – the worst off who benefit the most.

Case: Romas in Slovakia

Method

3 experiments

Study 1: Descriptive

-        Survey experiment – ca. 1000 respondents

-        Study 1a: how much money should to toward building a water well in a Slovak village or Roma settlement

-        Study 1b: helping regions in rural areas – with or without mentioning that Roma are cobeneficiaries.

 Study 2: Field experiment

-        Field experiment conducted in Slovak village with 50/50 ethnic Slovaks and Roma.

-        113 participants

-        Different frames

o   Control: Build social housing for Roma community with EU financing 95% of the cost

o   Equality: same allocation of money as for non-roma community

o   Reciprocity: Only Roma who help build the housing can live there

o   Need: Only the worst off Roma can live there.

Study 3: Generalization of results

-        Now a representative sample of Slovaks (1.000)

-        Same kinda treatment as in study 2.

-        Y1: I agree

-        Y2: Most people agree

Results

Study 1:

-        Mentioning Roma reduces support substantially.

Study 2:

-        Slovak

o   Equality and reciprocity gives significant HIGHER levels of support than control

-        Roma

o   Equality and reciprocity gives significant LOWER levels of support than control.

Study 3:

-        Only reciprocity yielded significant higher levels of support.

-        Personal support where generally higher than the perceived majority support.

Thoughts

Generalisability

-        Romas is a specific case with specific stereotypes. Perhaps especially dependent on “they don’t work only hustle

-        The field experiment environment is quite a unique setting – village with a 50/50 distribution.

-        Could be interesting to see if this holds with in-group?

Schäfer & Schwander, 2019

d

Income inequality has a negative effect on voter turnout = participation gap.  Does some different observational studies – some at a macro-level (OECD) and some at a state-level (Germany).  #income inequality

RQ

How does income equality negatively affect voter turnout?

Motivation:

Decreasing voter turnout = participation gap.

Especially among lower income etc.

Expectations

Contrary stances in the literature:

-        Positive relation between income equality and voter turnout

o   The poor mobilizes to ‘soak the rich’

o   The rich mobilize to prevent this

-        Negative relation between income equality and voter turn out

o   The poor gives up on voting because they have learned that the system is biased against them.

o   The rich seeks other ways to influence politics, e.g. lobbyism, donations.

Method

1.      Cross sectional for 21 OECD countries from 1980-2014

a.      They do a lot of robustness checks to enhance internal validity

                                                    i.     E.g. different standard errors and models that just varies the strength of the relation (however, one model yield insignificant relationship)

2.      German case with data from different districts (also over time)

a.      Also controls for stuff

3.      Who abstains: Election data: How does inequality affect vote choice in the 21 OECD countries.

Results

Trend

-        Study 1: Negative relation, moderate strength

1.      Study 2: Regional inequality reducer voter turnout. Seems to increase over time.

Who abstains?

-        Income inequality reduces voter turnout, especially among lower income groups

-        = The distance in voter turnout between income groups grows, the higher the income inequality there is.

Thoughts

A pretty fucking solid OLS – test with different measures, models and datasets.

The implications are pretty bad

The mechanisms need research, they say. Relation to the rise of populism?

Incerti, 2014

How to get renters to participate in local politics about housing

Problem: Homeowners more often than renters participate in town meetings about housing à get’s more influence Reminders to renters making a) participations cost lower, b) economic interest and c) cost of abstention, makes renters more likely to participate – either by written or vocal comments. Field experiment, US  #NIMBYism

Hwang, Hankinson and brown

Inequality in conditions for loans

Basically, arguing that banks during the housing crisis give worse conditions for loans (subprime loans) in minority neighbourhood in segregated metropolitan areas (high segregation) than in less segregated areas. So, if you are black and live in a black neighbourhood you are more likely to get a shitty loan than if you are black and live in a white/mixed neighbourhood. Everyone in these clusters areas, are vulnerable to these shitty loans. Observational, ols, US #housing/inequality

RQ

 

Expectations

H1: Subprime loans will be more concentrated in minority census tracts in highly segregated metropolitan areas than in less segregated metro­politan areas.

H2: Subprime loans will be more concentrated in clusters of minority census tracts in highly segregated metropolitan areas than in less segre­gated metropolitan areas.

 H3: Rates of subprime lending will be higher in minority census tracts in highly segregated metropolitan areas than in less segregated metro­politan areas.

Method

27% of the tracts where minority (over 40-50% of the population is black or Hispanic (Asian doesn’t count).

TI = 1 if only the minority clusters have subprime loans.

Clusters of subprime lending is overlapping a lot with the minority clusters

Some kind of OLS model with a lot of relevant controls

Results

Weak support for h1 (p < 0.10) – an insignificant decrease at like under 2 percentage points

Strong support for H2 and h3

Thoughts

Important implication: inequality in loans (and some inbuild racism?) à more inequality in wealth

Generalisability?

I don’t really understand the difference between h1 and h2 (and h3 for that matter) but I guess if you only look at tracts, there is no relationship between segregation and subprime lending. Because like if you live in a minority tract among a lot of majority tracts, you still get better loans. However, if you live in a minority tract amongst a lot of other minority tracts, you will get shitty loans. à thus, just look at a bigger level. Using tracts can seem a bit random, especially if your emphasize the fact that it is clusters of tracts.

Thus, their contribution to the literature is quite weak, I would say.

If not just a OLS, then it could have been made as a dif’n’dif or some matching. Make a control group of same income etc characteristic area and match it with the minority.

Einstein, Palmer and Glick, 2019

Confirms intuition – Highly screwed participation in local politics: Participants are likely to be older, male, longtime residents, voters in local elections and homeowners – and all opposing housing construction (NIMBYism. observational, descriptive aim, US #NIMBYIism

RQ

Who participates in local politics concerning housing and how do they participate.

Motivation

Politicians use town meetings etc. to form a view of their voters wishes.

Most literature on the subject is not observational – reason to do some descriptive research.

Some designs to increase participation actually increase socioeconomic inequalities in turnout (outcome or voter turnout I do not know).

A lot of American local level thing have restricted housing developers by empowering neighbourhood and environmental interest – however, participation opportunities might amplify some voices more than others.

Expectations

Expects anti new housing to show up, while pro new housing are less likely to show up

Method

Observational data (in contrast to most literature on the subject being survey data).

2015-2017

All citizen participations in planning board and zoning board meeting between 2015-2017 in 97 cities and towns in metropolitan Boston, Massachusetts.

Results

Participants are more likely to be older, male, longtime residents, voters in local elections and homeowners. All “overwhelmingly” opposed to the construction of new housing. But only hearings concerning the construction of more than one unit of housing.

Because these meetings includes the name and address of participants, they can couple it with some voting data, offering demographic data + voter turnout et.c.

Making a ‘balance tabel’ and see the statistical difference of mean commenters and mean non commenters. Also some kind of logged regeression

Those the participants are a unrepresentative group, disproportionately representing anti-housing vires.

Abou-Chadi, Cohen and Kurer, 2024

Increases in house rents in local area increases support for radical right party. Mechanism: relative decline in social status and pocket book. Explains why radical right parties gains support in some urban areas. Random effects, Germany #housing affordability, RRP

RQ

How chances in housing rent affect support for radical right.

Rental market risk describes individuals’ exposure to the latent economic threat from a rent price appreciation in their local environment.

Why radical right gets support in some urban areas à it is not only about periphery and globalisation.

Motivation

 

Mechanism

Primarily among those with lowest income. Why? Well because for those with low income, rent takes up a high share of their total income, making rising rents a bigger threat to their individual economy. Low income à higher rental market risk.

 

Social and economic threats à feeling discontent and marginalized à radical right appeals to these emotions à increased support for radical right.

The longer you have lived a place the more rental market risk. Thus a general expectation that these effects will be to find among long-term residents. Also, because you have to live one place a long time to experience status anxiety.

They lowkey argues against pocketbook effect (actually experiencing rents increases) and for geotropic effects (netto rent increases, thus perceived general increase in rents in you area (gentrification).

Expectations

H1: Increases in local market rents increase the probability of supporting the radical right among low-income renters.

H2: The positive effect of local markets rents on the probability of supporting the radical right among low-income renters is strongest in urban areas.

Stronger effect among long-termed residents.

Method

Data

-        27 million lejeannoncer aggregated to rent at a post code level.

-        8200 post codes in Germany

-        2005-2018

-        Combined with some individual level data. I think it is panel data

-        They also have some similar data on owners to compare with.

ONLY includes individual level data from tenants living there for more than 5 years (long-term residents).

Germany

Random effects instead of fixed effects

Results

The lower the income, the bigger the positive impact of local market rents on support for AFD. No effect of income on homeowners on the relation.

 

Test of mechanism

-        No effect of ‘hypothetical’ increases in rent (per square meter), across income groups.

-        However, there is a significant effect for net of local household rents.

So they write that perhaps a more direct

-         

 

Thoughts

Class

We expect people to respond to increasing house prices, but how?

This is about the market itself. The pocketbook effect, on the contrary, is about how your pocketbook is affected. This is more about a latent risk. Even though you are not directly affected by housing prices (rent loft etc.), they are responding to the market.

Ansell et al. 2022

Increases in house prices increases support for populist right parties. Mechanism: pocketbook, geotropic and mobility Support is strongest among homeowners (but also effect among renters = geotropic), not driven by labour market risk, immigration etc. Dif’n’dif, Nordic countries #Housing affordability, RRP

Research Question:
The study investigates whether changes in house prices influence local support for populist right-wing parties.

Specifically, it examines whether support is higher in areas with lower house price growth and weaker in areas with high house price growth in Nordic countries.

Motivation:
Wealth inequality is higher than income inequality.

Just looking at income inequality cannot explain it – median income is kinda the same in areas with low housing price growth as in areas with high housing price growth.

It is more about the cultural dimension than just the economic dimension.

Fluctuations in house prices. Geographical inequality in house prices.

 

Expectations/Hypothesis:
Higher support for populist right parties in areas with relatively low house price growth and vice versa.

Effect is more pronounced among homeowners than renters, but present among both groups.

Mechanisms:

·        Pocketbook effects – Homeowners experiencing declining property values become dissatisfied with mainstream parties.

·        Geotropic effects – Declining house prices signal broader community decline, increasing local resentment.

·        Mobility effects – Homeowners in areas with falling house prices feel trapped, reinforcing dissatisfaction with the political status quo.

Data:
The study uses registry data from Denmark, including detailed precinct-level voting patterns and house prices. It supplements this with municipality-level data from Sweden, Norway, and Finland to assess broader regional trends.

Method:
The researchers employ generalized difference-in-differences models to analyze the relationship between changes in house prices and shifts in populist party support. They also incorporate geo-coded survey data to assess individual-level responses.

Results:
The study finds strong evidence that declining house prices are associated with increased support for populist parties. The effect is particularly pronounced among homeowners (however not a clear relationship between house prices and renters support) and is not driven by labor market conditions, demographic changes, or migration patterns.

Weaker effect in Norway – perhaps because their populist party is not that anti-establishment because they have been present since 1973 ish.

Overall Conclusion:
The paper demonstrates that house price dynamics significantly impact political behavior, fueling populist support in declining areas. This suggests that economic geography plays a crucial role in shaping contemporary political movements.

Thoughts

Cool

They argue that similar effect should be present in other European countries – perhaps to greater extent as the percentage of homeowners generally is higher.

Bolet, Green and Gonzales, 2024

Environmental policies can gain more support in affected areas (coal communities) if 1) includes redistributive schemes to emendate the consequences of the climate policy – primarily if it benefits the whole community and 2) labour unions etc. are included in negotiations = process matters! Dif’n’'dif, Spain #Green transistion, fairness, compensation policies

RQ

Can ‘just transition’ strategies (JTS) affect support for government in in areas where the cost of green transition is highest.

JTS = climate policies combined with redistributive schemes to support affected workers and communities, with affected unions etc. included in negotiations.

Motivation

Green transitions imply diffuse benefits and concentrated cost. When CO2-heavy industries shuts down, the cost for the individual owner and community is high.

In the areas where cost is concentrated (e.g. coal mine areas), green transition often results in lower support for the government and the ‘green transition’-policies.

Few studies have examined if ‘just transition’ strategies can increase support for the government in these high-cost areas.

JTS = climate policies combined with redistributive schemes to support affected individuals and communities.

Mechanisms

JTS combines distributive justice and procedural justice.

Redistributive justice

·        direct payment – workers get compensated such as income support, early access to pensions or help to be retrained into a new profession.

·        Community-level compensation – such as placing another low-carbon industry in those areas.

·        However, studies indicates that community-level compensation is more popular among workers as it is perceived less as a bribe (bridge to geotropic effects)

Procedural justice

·        How political decisions is made matters.

·        Involving local actors etc. can increase support.

·        Thus the negotiations including labour unions is important because unions can also be a important player in mobilizing against a policy and they can have credibility and signal approval to voters.

Method

Data: Coal mine closure in Spain

Dif’n’dif and some interviews.

Municipalities with closing mines vs. municipalities without closing mines à argument: Random whether one municipality got coal is random.

Results

JTS increase support for government in general elections

Big effect as they find HIGHER support in the coalmining municipalities than in the non-coalmining municipalities (almost 2 percentage points.

What about the mechanisms:

·        Higher union density à higher support.

·        Coal-mine worker density has no significant effect on government support à indicates that the support is not driven by direct payment mechanisms (arguably).

·        Interview data and other literature suggest that coal-miners prefer community-level redistribution.

Thoughts

Why does this matter

·        Showing that compensative policies can mitigate the negative effects on incumbent support when implementing a green strategy.

·        we don’t know their preference for clime policies, but they bear the burden of green transition. a good example of economic vulnarbility translating into policy preference, the government has a important role in mitigating this effect kinda.

Generalisability

·        geotropic effects (community level) are shown to matter elsewhere (Ansell et al regarding housing prices and support for RRP)

·        we believe that this can explaprolate because of the community wide benefit – makes laura more likely to think that people feel more seen and protetcted by the government – move beyond ‘I loose my job, I get a new job’ to other examples that is not coupled with jobs but perhaps generalise to nature catastrophes etc.

·        perhaps less generalisability regarding labour union

·        what about Denmark and farming?

 

Voeten, 2024

Increases in household energy cost after a climate policy increases support for radical right parties among energy sensitive renters (those with separate utility bills and those who use a lot of their income on energy). Dif’n’dif, Netherland #housing affordability, rrp

RQ

Does increases in household energy cost (after a implementation of a new climate policy) increase support for RRP. Especially, among energy sensitive renters (those who pay for energy separately from their rent and the energy poor).

Case: The Dutch policy changes that increased taxes on household natural gas consumption and redistributed the revenues as subsidies for installing renewable energy sources (solceller på taget).

Motivation

Economic voting (in terms of climate policies often being disproportionately more though on lower income (at least in regards to the Dutch case) might actually play a bigger role in explaining increase support to RRP.

Expectations

Energy sensitive renters (e.g. those affected = separate utility bill and energy poor) are more likely to vote for a RRP party after a climate policy that increases household energy cost.

Mechanisms

Economic voting.

Renters with a separate utility bill are not likely to benefit of the polices redistributive schemes as their landlords don’t have an incentive to install solar panels etc.

RRP were the only vocal opponents of the gas.

Data

2007-2020

Method

Dif’n’dif between affected and unaffected citizens (gas or no gas)

Results

After the policy change, renters with separate utilities bills became more supportive of RRP and less sympathetic towards the green party. Smaller effect, however still significant, among the energy poor (people who spend 10% of their income on household energy or struggles in other ways to pay their energy bills).

However, these people did not change their views on immigration, the EU or changes their left-right placement à indicating that this is just a economic issue??.

Also, indicating that it goes the other way. Non-energy sensitive people became less sympathetic towards the RRP, implicating that IF the policy is very popular, it could decrease the support for RRP. However, the RRP can just adjust their stances, perhaps not being anti-green but just talk about the importance of redistributive schemes etc.

 

Thoughts

Class

A lot of rent control in European countries. Thus, energy cost might feel more urgent. At least in the German case, energy cost is higher in rural than urban areas à poses a risk for people that in relation to the whole renter/homeowner discrepancies, rural areas are more at risk.

These articles all talks to the fact when prices or the risk of prices going up, it affects voting behaviour. The takeaway here is that housing and energy has merged into two important components. Historically, most focus has been on the labour market, but now we are focusing on housing and energy as well.

However the mechanism behind the relation is really explored in these articles.

Scheiring et al. 2024

Meta studies of studies with a causal research design examining the relationship between economic insecurity (broadly defined) and support for populist parties (both left and right wing). Economic insecurity increases support for right wing populist parties, however the results for left wing populist parties are not clear. Perhaps because they aren’t as widespread. Some publication bias. #rrp

RQ

Do economic insecurity increase support for populistic parties

Motivation

Schering et al. (2024) wanna debunk those who says that economic insecurity has no relation in support for populist parties.

Describes the main explanations as globalisations changes in labour market (the left behinds) and cultural explanations (symbolic loss of social status).

Economic insecurity however is important = ‘risk of economic loss faced by workers and households as they encounter the unpredictable events of social life = both individual level but also macro level (foreign currency debt shocks).

Expectations

That economic insecurity increases support for populistic parties - both left and right wing populistic parties, even though right wing populistic parties are more common.

Data

Only studies with a quasi-experimental, field experiment or survey experiment that looks at the relationship between economic insecurity and support for left- or right-wing populistic parties.

Method

Meta study

144 studies included

Also looks at publication bias

Results

Economic insecurity increases support for populism – about one-third of recent surges in populism. Significant results for right wing populism, but not clear results concerning left wing populism.

Heterogeneous effect depending on the kind of economic insecurity, the most robust results stems from housing demand shocks, subjective economic hardship, austerity and foreign currency debt shocks.

Publication bias but still significant results after controlling for thus

Thoughts

Doesn’t go into the meachanisms and why’s

Motivation is not that convincing - like are anybody claiming that economic insecurity doesn’t broadly increase support for rrp

Hilbig and Vief, 2022

Rent control annuls NIMBY and induces YINMBY for those who live in gentrified areas and where rent control makes a substantial difference in the rent. Rent control reduces worries for displacement and gentrified.  RDD, Berlin #housing affordability, rrp

RQ

Rent control reduces NIMBYism (among renters) and induces YIMBY (yes-in-my-back-yard).

Motivation

Well, one could imagine that rent control has some effects.

NIMBY has a economic root caused by fear of gentrification and displacement.

Expectations/theory

Housing crisis

·        Demand side: Increased demand after housing because of population growth, increased per capita consumption of space and urbanization.

·        Supply side: Supply is slow, mainly because of hampering by zoning laws, unpopularity and construction laws made to entertain

NIMBY: refers to opposition to land use projects with spatially concentrated costs with diffuse benefits. Both opposition to new housing projects as well as new residents

Rent control reduces NIMBYism (among renters) and induces YIMBY (yes-in-my-back-yard).

Mechanisms

NIMBY is caused by economic fears – fears of displacement and gentrification. New construction in cities is usually target wealth individual, which increases the risk of house rents to increase because of gentrification etc. However, rent control (if it is with any substance, gives renters enough economic security to overcome this. Argues mechanism: Rent control gives people an “additional” income in some way, making them more suited to “stomach” the gentrification. Rent control also gives people more economic security because they know that their rent can increase uncontrollable.

However, their preregistered expectation was that rent control increases NIMBY, because they would have a small need for new construction, thus lower acceptance of the noise etc. building would cause. They are more inclined to leave a long time in their rent controlled homes, making them more inclined to become invested in neighborhood because of the noise etc and they will have the same worries as house owners (we don’t want some losers to move in or increased competition for ressources (schools, hospital and transport).

Data

Survey send to 32.000 respondents (respond rate at 9%)

Method

RDD because rent control in Berlin was only applied for those appartments build before 2014.

Results

Rent control induces YINMBY

·        People with rent control report improved financial security. Rent controls doesn’t help the housing crisis – people are more prone to live in the house longer. The results show that for existing residents, it actually relives worries about financial security.

·        No impact on people being less or more willing to move.

·        Rent control has no effect among those where tenants receive very small rent reductions = not a symbolic effect.

·        The effect is especially pronounced in areas with a high degree of gentrification. However, if there is low gentrification, no YIMBY is induced.

Thoughts

Quite important that this is only for renters, driven by people in gentrification areas where rent control made a substantial reduction of their rent.

Urban – perhaps this will work differently somewhere else, where gentrification is no real fear.

The mechanism proclaimed in other literature is often the fear from home owners that the construction will lower the value of their property.

Strong points

·        Preregistred hypothesis expecting more NIMBY because of rent control

Colombo et al, 2024

The literature claims the economic inequality reduces support for the political system, however, most people are not observant of national level economic inequality but are more aware of the inequality observed at a local level Even with a lot of fiddling with models and different measures of local-level economic inequality, there is no indication of local-level economic inequality reducing support for the political system. Draw backs ·        Is Denmark the right place to test this? ·        Maybe national level inequality is more important than the authors think. Fixed effects and cross sectional, DK and ESS #inequality

RQ

Does economic inequality reduce support for the political system?

Motivation

Widespread claim in the literature is that economic inequality reduces support for the political system among the disadvantaged.

However, Colombo et al. (2024) criticize this claim for assuming that people actually know how high the inequality is and how they fit in in that distribution. Thus, national-level inequality is a very bad measure. Instead peoples local experiences of economic inequality should matter instead.

Mechanisms

Economic inequality à Reduces support for political system

·        Signal that the democratic system does not deliver an equal distribution of resources, fostering the perception among the poorer segments of the country, that the system does not work for them. For whom the inequality is the biggest disadvantage, the more disapproving of that system.

Data

The study integrates three key data sources (all Danish respondents)

• Survey Data: Various Danish public opinion surveys from 1999 to 2017, including:

  • The European Social Survey (ESS)

  • The Social and Political Panel Study

• Danish Registry Data: Population-wide administrative data containing socioeconomic characteristics at the individual level.

• Geo-Referenced Data: They link individuals’ locations (place of residence) to economic data to measure local inequality.

Y = support for system

·        Regime performance (satisfaction with democracy)

·        Regime institutions (trust in institutions e.g. the police, legal system and parliament)

·        Trust in politicians

X:

·        Gini coefficient of the local INCOME distribution.

·        Gini coefficient of the local WEALTH distribution.

·        Gini coefficient of the local value of vars

Method

2-way fixed effects

And some cross section I think

Results

Makes a multiverse plot, where most of the coefficients are insignificant – as well as the coefficients in the cross-section and fixed effects regression output, where only 1 of the 6 coefficients was significant.

9/16 statsically significant coefficients in the multiverse model is positive, indicating the reverse effect of income inequality.

Interaction with individual income.

·        Distribution of the marginal coefficients

·        Figure 3: little to indicate that inequality matters differently depending on the individual income

= NO EFFECT OF ECONOMIC INEQUALITY

Discussion

The authors mentions that perhaps are national level measure of economic inequality more important than they thought. Hard to rule in, but just as hard (and perhaps harder) to rule out.

Thoughts

Good, extra strongpoint that they use multiple measures of economic inequality and shows that even if they fiddle a lot with the models, almost no significant results are shown, and if they are significant, a lot of the time they are positive, underscoring that there is no relation between local level economic inequality and support for the political system.

This is Denmark. We all think we are middle class, kinda. So not the generelisable. Least-likely case.

Weidemann, 2024

Robin hood paradox: Inequality is rising, why no more redistributive policies? My formulation, when economic inequality have spatial variation – with some constituencies having a high degree of income inequality and some low. Then supporters of redistribution are only concentrated in a small number of constituencies, giving small incentives for the parties to promote redistribution as part of the political campaign in plurality systems. Regression, OECD, UK #inequality

RQ

How does the spatial concentration of rising economic inequality affect support for redistributive policies in plurality voting

Motivation

Puzzle: ‘The Robin Hood’-paradox: Why has there not been an increasing support for left—wing parties/ redistributive polices as the economic inequality is rising? This should be expected in standard political economy models. However, empirically we often observe the opposite.

Expectations

‘Institutions matters’-argument

Geographically concentrated inequality undermines the political logic of redistribution (more inequality à more demand for redistribution) under plurality rule (winner-takes-the-all).

Mechanisms

In the case that economic inequality is spatially divided AND the median electorate district (median in relation to income equality) is lower than the national average, the link between income equality and support for redistribution will become undermined in plurality systems.

Wiedemann draws on theory on basic strategic party behaviour. In plurality system, the median voter in the median district is an important electoral target. Therefore left-wing parties will choose a more centrist and less distributive policies to appeal to the median voter in the median district (if the median district is as argued before, less inequal than the national average.

My formulation, when economic inequality have spatial variation – with some constituencies having a high degree of income inequality and some low. Then supporters of redistribution are only concentrated in a small number of constituencies, giving small incentives for the parties to promote redistribution as part of the political campaign in plurality systems.

Data

Cross- national data in OECD countries

Local-level data in the UK

Comparative Manfesto data

Method

 

Results

Cross-national analysis

·        Y = fiscal policies

·        Spatial inequality is associated with less redistribution in countries with plurality rule than in countries with proportionality rule.

UK analysis

·        Y = voting for labour and redistributive preferences

·        The spatial concentration of inequality makes the median constituency less unequal than the national average.

·        This limits the demand for redistribution and electoral support for left-wing parties only a few densely populated urban constituencies with high inequality

 

Weidemann (2024 argues that his findings offer an “explanation of why rising inequality has not led to more redistribution, which suggests that political geography can weaken political responses to inequality and electoral representation.”

 

Thoughts

Like Colombo et al. (2024), Wiedemann argues that one should look at more local level measures of economic inequality but with another spatial focus than Colombo et al.

Compare it with all the other articles we have read, claiming that economic inequality leads more support for the far right. Is it about institutions. I don’t think so.

I don’t understand this papers argument. There is to many ifs. However, the fundamental, that spatial concentration of high inequality in a few constituencies ‘undermines’ electoral representation of the voters preference for redistribution. But does it? Is it confounded by factors behind high inequality. That many low income people in the urban areas support right wing populist parties. I don’t know. I’m confused and I think this set up is hard to understand and seems a bit to planned.

Class

In this whole country, inequality is rising, but people are not responding by supporting redistribution.

Essentially, if a district is experiencing a 90% increase in equality (really struggling) and the other are only facing moderate or small increases in inequality, the average still goes up, however, it is distributed quite unequally.

Elkjær and Wlezien, 2024

To include or not to include ‘don’t know

To include ‘don’t know’ or not as a response category. You should include it, or people will just respond kinda randomly if they have low political knowledge, especially on low salience issue. On high salient issues to include or not to include doesn’t matter that much. Argument for not including: forced consideration Survey experiment where treatment in including DK or omitting DK.

Cavaille et al., 2024

Which measure most reflect opinion-congrent actions

Comparing different measures of support on their performance in measuring whether respondents are likely to translate opinion into opinion-congruent actions. ·        Likert (strenght of support) ·        Likert + (strenght of support + salience) ·        Quadratic voting in research surveys (QVRS) – some measure where people are shown a menu of policies in which they can buy votes for or against. The more votes one buys for/against one policy, the more the votes cost. Less salient issues = favour of QVRS Y = opinion-congruent actions: donating to a non-profit advocating gun control etc.

There is a difference between expressing support/oppose a policy and actually translate this support/oppose into opinion-congruent action e.g. turn up to vote. Chavaille et al. therefore sets out to compare different measures of policy support.

Motivation

There is a difference between expressing support/oppose a policy and actually translate this support/oppose into opinion-congruent action e.g. turn up to vote. Chavaille et al. therefore sets out to compare different measures of policy support.

What do they do

They test 3 different measures:

Likert

·        Likert measures the strength of support/opposition to a policy – here on a 3-point scale (weird)

Likert +

·        Strength of support/opposition to a policy + whether the issue is PERSONALLY IMPORTANT to them measured on a 5-point scale.

Quadratic voting

·        In quadratic voting, respondents are given a fixed budget to ‘buy’ votes in favor og against policies. The cost of votes is quadratic – 1 vote for/against = 1, 2 votes for/against = 4, 3 votes for/against = 9, 4 votes for/against = 16 etc.

·        In this paper, respondents are given a total of 100 points, making it possible to buy a total number of 10 votes for/against a policy. Respondents are not forces to buy votes.

Theoretical argument

I guess they argue that QVRS are best at measuring whether people actually are willing to translate their opinion to action, thus enhancing measurement validity.

Method

Survey experiment

Independent:

·         treatment is random assignment to the 3 different measurement techniques.

 Dependent:

·        Different task indicating opinion-congruent action such as donating to a non-profit advocating for gun control or letter written to a senator about a minimum wage bill.

·        Not with their own money

Data

 

Results

They test the measurements performance on three parameters:

1.      maximizing interpretable variation,

2.      minimizing survey costs,

3.      minimizing noise.

QVRS performs well at 1, a bit worse on 2 because the surveys is longer and higher drop out, and the improvement on 3 is minimal.

For high politicised issue, variance is lower in Likert than in Likert+, speaks in favour of QVRS.

For less politicised issues, less difference.

Relation to other readings and is the text convincing

The trade off thing – responses change when the trade off between policies are made clear.

A bit difficult to measure ‘opinion-congruent actions’ in the setting of a surveys.