4. Democracy and Median Voter Theorem

Median Voter Theorem: Equilibrium

• parties choose the Median voter ideal policy point

• median voter: half voters prefer lower policy point, half voters prefer higher policy point

• median voter theorem: under certain assumptions, majority voting yields the policy outcome preferred by the median voter

Business analogy: Hotelling’s Model

• consumers distributed uniformly along a street

• 2 ice cream trucks sell same ice cream for same price

• consumers buy from closest truck

• result: ice cream trucks converge to the middle of street

Median Voter model: Assumptions

1. single dimensional voting and single-peaked preferences

   • voters care about single issue

   • voters have bliss point and utility falls as you move away from that point

2. Two candidate or parties

3. no party ideology — parties only care about winning office

   • if they did care, median voter won’t hold bc. too se on personal beliefs

4. no selective voting — everyone votes

5. no money in politics

6. full information — politicians know voter preference, voters know party positions

what if we had multidimensional policy space?

problem because there is no clear equilibrium

if there were more (3) would parties have incentive to move away from MV policy position?

deviate because parties would have an incentive to move away bc they will ensure a win

What happened to Britain’s Median Voter?

• 1997 labour thought they needed to cater to mondeo man and the worcester woman

• Tony bliar noticed that labors who had done well were now becoming tory and viewed labor as hindering him

• convergence to the median depends on which issues are more salient

  • economic issues = parties closer

  • other issues like Brexit = ideological stance

• mvt may hold within political groups — leave/remain in EU supporting parties

• median voter can change over time and does not have to be in the middle of political spectrum

• no full info: different parties can have different veliefs on median voter

• 2019 conservaties bet on MV being working class, leaving voting northern

evidence on MVT: how can we provide evidence on whether the median voter preferences rather than party/politician preferences determine policy?

• having to different parties means the two parties are not equal and therefore have different median voters in each so against the MVT

• to have MVT you need same median voter and same policy

Close races

• we will see that MVT is still close even through different outcome

• if you compare D1 and D2 they should vote the same, because same MV same policy

MVT: Evidence from Lee, Moretti, and Butler

• shows when republicans barely won and when democrats barely won

• provides evidence against MVT using regression D.

• suggests party affiliation strongly determines policy outcomes rather than convergence to the MV preference

MVT: Evidence from gerber and Lewis

• analyze voting record and legislator behavior in LA in 1990s

• MV preference are better predictor of legislative behavior in homogenous than heterogenous districts

• homogenous districts = more MV influence — voters have similar policy preferences, legislator will align with to get reelected

• Heterogenous = more party influence — voters have polarized/different views, legislators will rely on party platform

• big picture: whether there is convergence to the median depends on the context, MVT does not always predict legislator behavior

Lecture 4 Summary

• MVT: majority voting yields the policy outcome preferred by the median voter

  • requires certain assumptions that may not hold

• evidence on party policy convergence is mixed: depends on setting