LT Week 9 Flashcards
The value of democracy & a free media
Motivation
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
To quantify the magnitude of ethnic favouritism in road investment and what extent this is affected by democratisation.
Special settings
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
Ethnic favouritisim is where those of the same ethnicity of the controlling government benefit from patronage and public policy decisions so receive disproportionate share of public resources
Can arise when weak political institutions are unable to constrain governments from discriminating among citizens; important to understand political and economic performance to discover extent if/of ethnic favouritism or if emergence of democracy has helped mitigate it
Theory
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
“Quantify extent of ethnic favouritism in public resource allocation in post-independence Africa
* Challenge: to understand spending by group due to poor statistical offices
* Examine whether the transition in/out of democracy affects extent of ethnic favouritism
* Challenge: often long tenures of governments making difficult to observe changes in power on ethnic groups
* Challenge: need to observe switch between democracy and autocracy under same leader which is uncommon
Use road building in Kenyan districts due to high ethnic segregation across districts and road expenditure can be directly measured; post-independence allows for transitions between autocracy and democreacy under several leaders to assess role of democracy. Kenya is a strong proxy for rest of sub-Saharan Africa during this time.”
Empirical design
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
“Create model of presidential public resource allocation across districts showing degree of ethnic favouritism determined by constraints on executive action characterised by different political regimes
Construct counterfactual road network
Approach 1: graphical analysis to examine ratio of district’s share of road spending versus population share
Approach 2: use regression to estimate road spending capturing coethnicity with president using indicator value (1 where majority ethnicity share) and indicator variable for multi-party democracy. Control for geographical/demographic/economic variables using baseline vector that might impact road spending/construction to control for where it might take place; also controls for district and year fixed effects”
Data
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
Construct district level panel data on road expenditure on 41 Kenyan districs using historical maps and cross-check the district road expenditure data with district road construction data
Key findings
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
- Co-ethnic districts receive greater road expenditure under autocracy
- Road construction decisions were not based on market potential
- Democracy dampens favouritism
- Ethnicity of other officials doesn’t matter,
perhaps except VP
“Kenyan districts sharing ethnicity as president receive 2x expnditure on roads and 5x length of paved roads as would be predicted by population share –> inconsistent across regime (biases disappear during democracy)
Counterfactual road network shows no evidence of ethnic favouritism or due to regime –> effects in actual data not due to coethnic disticts just happening to have higher market share and regime effect not a coinceidence with natural expansion of roads”
Interpretation / policy implications
The Value of Democracy: Evidence from Road Building in Kenya (Burgess et al., 2015)
“Shows that even in imperfect democratic institutions there is vaue in limiting power of executive
Not causal evidence by reason for further research”
Motivation
Weigel (2020): The Participation Dividend of Taxation
Gives important insight into the question on whether taxing people makes them more politically active.
Special settings
Weigel (2020): The Participation Dividend of Taxation
“This article provides evidence from a fragile state that citizens demand more of a voice in the government when it tries to tax them. To do so, it examines a field experiment (RCT) randomizing property tax collection across 356 neighbourhoods of a large Congolese city. Kananga is a city of roughly 1 million. Capital of Kasï Central province. In 2015, total provincial tax revenues amounted to a paltry $0.23 per person in the province. Prior to 2016, property owners were in theory supposed to visit the tax ministry themselves to pay.”
Theory
Weigel (2020): The Participation Dividend of Taxation
“Why would someone pay taxes if they do not have to?
- Entitlement: taxpayer expect reciprocal benefits from paying taxes e.g. public good provision. This owuld be consistent with a finding that showed households who pay tax are the ones who participate more
- Bargaining: tax soliciation provides citizens with ‘bargaining chip’: demand public goods in exchange for future compliance. Able to hold government to account if you are paying them
- Updating: Changes in the way one views the enforcement capability of the government. Tax collection send a signal of state capacity, which causes citizens to update their beliefs about the government and expect higher returns to participation”
Empirical design
Weigel (2020): The Participation Dividend of Taxation
“Experimental roll out of property tax: The treatment, randomly assigned on the neighbourhood level, is door-to-door property tax collection campaign which ran from April to December 2016. Among 431 neighbourhoods 253 were selected randomly to receive the tax campaign in the first phase. Control scheduled to receive the tax campaign in mid-2017. Distribution of fliers in treatment and control regions of an informational flier that let people know of tax collectors beginning door-to-door collection to ‘secure the province, to kickstart economic development’
This article examines three set of outcomes:
1.) Did the experiment even work?
- Visited by collector: an indicator that the household received visits from provincial tax collectors in 2016
- Registered as taxpayer: an indicator that the household was registered by collectors and assigned a unique tax ID. This is measured by the presence of a tax ID on the door or wall of a house
- Property tax compliance: an indicator for verified payment of the property tax in 2016, on the household or neighbourhood level
2.) Second set of outcomes concerns political engagement. In early 2017 the provincial government held five town hall meetings. Chaired by the finance minister and the director general of the tax ministry, the meetings sought to promote dialog between officials and citizens about taxations and public spending in Kananga Other measure of participation is the submission of anonymous evaluations of the provincial government to a locked drop box in downtown Kananga and beliefs about government “
Data
Weigel (2020): The Participation Dividend of Taxation
“Data come from four sources: (i) administrative data on property tax payment, (ii) a baseline survey before the campaign, (iii) a midline survey during the campaign, and (iv) an endline survey
after the campaign.”
Key findings
Weigel (2020): The Participation Dividend of Taxation
“1. Campaign increases tax visits, registration and compliance (the RCT works essentially - all of this in Table III)
- 81.5 percentage point increase (be careful of units as we have dependent binary variable) in tax collection within the treatment group
- Tax compliance dependent variable shows that in the control group tax compliance is basically non existent (0.06% of households pay taxes). However, we see a 10 percentage point increase in tax compliance in the treatment but still very much the miority (90% of population still not complying)
- Column 5: 367.295 CF increase in tax revenue per person very significant
- Campaign increases political participation
- Column 1: town hall meeting attendance post campaign increases by 4.5 percentage points
- Column 2: evaluation form submission increases by 2.4 percentage points - Campaign impacts beliefs on state capacity
- Campaign inreased people’s beliefs that the government is responsible for public goods provision, increases people’s belief in the enforcement and ability of the state when it came to tax collection (these effects on state’s information on tax payers and compliance are strongest) - Table VI is all about the potential mechanisms whereby the roll out of this RCT (look at theory section for context here)
- Looks only at the treatment group here and looks at whether people participating in politics on whether they are a registered taxpayer/comply with property tax. i.e. are the ones actually paying the tax, the ones that are participating in political sphere
- Do not really need to worry about OLS and IV, just look at how columns 7 and 8 show that while registering as a taxpayer does make you more politically engaged, complying as a taxpayer does not make you more likely to be politicaly active
- Conclusion: not a story of individual entitlement
- Paper also rules out updating as a potential mechanism by looking at sillovers in the control group (asks whether people who live closer to treatment within control group more likely to be politically engaged and the answer is no)
- As such the paper argues that the bargaining mechanism is strongest - descriptive/qualitative evidence on nature of town hall meetings suggest demands for better services
“
Interpretation / policy implications
Weigel (2020): The Participation Dividend of Taxation
This paper is really impressive in exploring the effects of tax policy and political engagement. With state capacity being veyr hard to quantify, tax collection abilities is a great proxy for it. The ability for a state to tax is significant for development and this paper gives great insight into the mechanisms whereby the state is held to account.
how to improve
its legitimacy
- develop a multi-party democracy?
- build a free and independent press, targeting all groups of society?
- increase the checks & balances on the executive?
- decentralise power and decision making?
- set quotas for various ethnicities to be represented in government?
- reward officials if they achieve specified targets?
