rafaelchduran
commented
3 years ago Rewriting of Reelection Backfire considering Pablo´s and Hye Young´s comments
This revision needs to (a) make a linear argument and (b) erase all the noise. What is the paper really about? Delegation is a political problem. Thus leave aside:
- principal-agent discussions
- IR literature
What to include that is missing? -thing on the context of what citizen are expecting and how they like Hawks. -explain why do citizens still vote for the incumben if violence increases. Speak about the short to long term credit claiming game politicians play in Mexico.
On style
- [x] don't use coloquial phrases
- [x] have 5-7 sections snot 12.
- [x] cut down no relevant papers. Hye Young gave an example with delegation to
- [x] check typos and also the bibliography of Caballero 2015.
- [x] do not use the term local executives but mayors
On small things
- [x] Motolinia paper might not be negative consequences.
- [x] Decide which literature you want to appeal to.
- [x] “Adds color to the conclusion” : My finding add nuances to the existing literature. Degree of delegation may be the difference.
On introduction: contract of what the reader is to expect on the paper
-
[x] Use the following structure:
- Clarify the concept of delegation: Mayors are less likely to sign the form and give less authority to governors to take care of public security.
- Big motivation. 1 paragraph on the puzzle of delegation in politics
- This paper studies... the role of reelection incentives on delegation. Reelection incentives puts more pressure on incumbents up for reelection to signal competence to voters and this may lead not to delegate even when its desirable to do so. You want to signal competence. You do so by owning rather than delegation policy decisions to others. Mayors believe that from those signals of who is in charge of policy voters can learn that he is in charge when he runs from reelection.
- I explore this in a setting where mayors mayors face a dilemma to delegate through signing security cooperation agreements. Talk about the signing as a measure of delegation and not cooperation.
- How do I do it? I leverage the staggered implementation of the 2014 Reform and test the effect on the likelihood of signing. I find that .... main finding. Moreover I find the differential effect according to the importance voters attach to security and trafficking as the main concern, alignment.
- Summary of findings if not done before
- Summary of robustness checks. Drop falsification test.
- 8 and 9. Contribution to the literature
-
[x] Describe well what delegation implies in this paper and its not cooperation. Describe the centralized commands. Pablo says that ``This is very important, and this is what should receive the most attention. You need to make the case that this is indeed delegation, and not cooperation. Or that it substantially decreases the ability of local mayors to claim credit for security policy."
-
[x] check the delegation number according to my data not the one from the National Census.
On theory What do I want to say?
- Delegation in some contexts like security policy is the most efficient thing to do. Its not obvious because of spillovers and heterogeneity of tastes though. 1.1 Most delegation decisions are made considering policy product. 1.2 But its a hard choice because of spillover and het tastes. 1.3 But in security spillovers are greater than heterogeneity and need a lot of coordination and resources to be tackled. The large ballon effect literature shows this.
- But there is a non-obvious choice for politicians up for reelection. The puzzle. (Fix specially the paragraph in page 12). Having more electoral pressure through reelection leads mayors to signal. You want to signal competence. You do so by owning rather than delegation policy decisions to others. State the novelty that electoral concerns matters for delegation. 2.1 Incumbents need to believe that signals of competence to voters increases there reelection chances. 2.2 Incumbents believe that taking charge of policy sends a signal to voters about their competence. Talk about the benefits that they get from taking charge of policy instead of delegating it to the governor. 2.3 They take charge of policy by not signing security cooperation agreements with governors. 2.4 Voters wants more security. But if delegate they cannot gain the electoral gain. So in some sense reelection concern . Clarify that its that they prefer haws and that they respond to delegation. 2.5 Need a paragraph that shows why reelection incentives affect delegation decision. So instead of the one on three reasons.
- Credit claiming and responsiveness. Say that if we only see the former we only see the act of signing but not the act of persecuting crime.
- When is the more salient? A) when citizens value a policy and B) when not aligned. 4.1 For this story to be credible citizens need to value the policy. An asymmetry is expected according to the level of concern voters hold. 4.2 For this story to be credible citizens need to be able to identify the responsible of policy. They do so when candidates are aligned.
There is a comment from Pablo that I don't understand. He says that the theory is not about differentiation but rather about fully controlling policy (i.e. taking the bull by the horn) vs. not doing so. If not I need to adjust for the theory. I believe this is related to Hye Young´s point that I need to differentiate between effort and credit claiming only, i.e. between an interest of controlling policy or showing only that you intend to do something but don't do it for real. I can use expenses in security to show that this is the case.
On research design
- [x] explain thoroughly the placebo with other political players in the data section. Don't mention it on the introduction. Vertical competition in public goods to explain on the falsification test. To show why this is a placebo test.
- [x] on the data describe better what is effort and how to measure it. Also say that I will use expenses on public security as measures of effort. I have the same problem all the conflict literature (and Elan) on having anti narcotic activity as a measure of both effort and violence. I could claim that conditional on violence its a measure of effort but might get pushback. So add the expenses.
- [x] acknowledge that pressure to sign agreement comes from the party. Outside forces like the party or the president might explain the delegation behavior. But these differences are ruled out by the design.
On results
- [x] show the naive diff in diff results
- [x] differentiate the robustness of the main result using different methods and the ones to test the theory. I should separate these into different subsections
- [x] explain placebos really well in the preferences section.
- [x] explain well the measure of preferences
- [x] acknowledge that I do not find evidence on insecurity and punishing criminals
- [x] Redo Figure 6. Drop the Total interaction effect and establish which ones are placebos.
- [x] Redo Figure 7. Drop the alignment with PRI.
- [x] do not talk about unintended consequences but just consequences overall
- [ ] run interactions by winning margin pre treatment
On effort and outcome: the disconnection between effort and competence
- Maybe there is a difference between effort and output. They talk and try to hire more police, dispatch more police, but the outcome doesn’t change. What voters observe is the effort so that politician. Distinguish effort from output. Accountability is the actual policy outcome.
- Need to distinguish between effort and outcomes. If voters observe effort and they would reward you. If they care about the outcome, so why would they reward you.
On conclusions`
- [ ] acknowledge that there are other potential stories like state capture or differences in the quality of politicians that could explain the results.
- [x] acknowledge that it seems that its credit claiming and differentiation but not taking charge of policy and taking the bull by the horns.
- [x] acknowledge that pressure to sign agreement comes from the party. Outside forces like the party or the president might explain the delegation behavior. But these differences are ruled out by the design.
- [x] drop the IV
- [x] Acknowledge Jane Esberg's story: You might get to this, but there is a more nefarious read of this, which is that incumbents up for reelection in areas of mexico with high levels of narcotrafficking want the capacity to flexibly negotiate terms with criminal groups.
- [ ] By delegating, mayors also loose security forces as political brokers. Example of Mexico's intermediate elections.
Todos