AER decision // Submit to AEJ

[santiagohermo]

In this issue I will draft and send a letter to the AER to ask about the status of our submission.

EDIT: After hearing back from the AER, we will submit to an AEJ journal. We will prepare a cover letter and revise it in this issue.

[santiagohermo]

Sent an email through manuscriptcentral

Dear Alexandra Wells:

We hope this email finds you well. We are writing to follow up on our paper, "Minimum Wage as a Place-Based Policy: Evidence from US Housing Rental Markets," which we submitted to the AER for possible publication on August 22, 2022. We would be grateful if you could provide any information on its status.

We understand that the reviewing process is time-consuming and appreciate the effort that goes into making a decision. We apologize for any inconvenience caused by our eagerness to find out the paper's status. Once again, thank you for considering our submission.

Sincerely, Our names

I will leave issue open waiting for a reply.

[santiagohermo]

I added a submission page to the wiki with the AER decision @diegogentilepassaro @gabrieleborg. Looking forward to talking tomorrow!

[santiagohermo]

Guidelines to submit to the AEJ Policy: https://www.aeaweb.org/journals/pol/submissions/guidelines#:~:text=All%20manuscripts%20should%20be%20submitted,submissions%20may%20be%20summarily%20rejected. (See section on papers previously submitted to the AER)

[diegogentilepassaro]

Adding that this is true for most AER journals, see for example AEJ Applied:

https://www.aeaweb.org/journals/app/submissions/guidelines

[santiagohermo]

Thoughts following phone call with @diegogentilepassaro (FYI @gabrieleborg):

• Try to submit to a new journal quickly. In particular, we would like to take the short route for an AEJ.

• To do this, we need to prepare a cover letter specifying what we would change for the AEJ editor.

  • If the editor accepts, then we would have a few months to implement what we promised.
  • We would split tasks between @diegogentilepassaro @gabrieleborg and @santiagohermo

  So, in the short run we want to:

• Re-read carefully the reports and make a list with comments. Mark the ones that we would and wouldn't implement.

• Make sure we include something from each report.

• Meet to discuss.

• Prepare cover letter to AEJ editor.

[santiagohermo]

I will start summarizing comments from the reports here @diegogentilepassaro @gabrieleborg. Feel free to edit this comment!

Comments on referee reports

Referee 1

1. Argues that adjustment of rents should be sluggish. "The fact that the results are quite senstive to time frame is not good news for the credibility of results."
2. Argues that selection into listing may be driving the results.
3. ZIP codes with missing rents suggests problem with selection highlighted in 2 may be worse.
4. Suggests using survey data: PSID, CEX (used in Aaronson et al 2012, which btw we hadn't looked into before). Asks us to compare with that paper.
5. Says that the estimates are too large with some quick computation (we should reviste it carefully). Asks for back-of-the-envelope calculation.
6. Says that our estimates of the effect of income are too large relative to Cengiz et al 2019, when we explicitly compare with them and argue they are consistent.
7. Didn't like the fact that we "restrict attention to one type of workers" in the model. Thinks we should discuss household structure more.

Referee 2

Positive referee!

1. Would like to see the effect of the minimum wage on local prices. Is magnitude consistent?
   • Data from Diamond & Moretti (2022)?
2. Accounting exercise, are estimates plausible?
   • "The goal here is to see whether the paper’s estimates are 'ballpark plausible.' "
3. Wage spillovers to contiguous jurisdictions
   • "If there are large wage spillover effects on contiguous jurisdictions, then an increase in the workplace minimum wage on wages relative to contiguous jurisdictions should not have a significant effect on housing demand in the relevant zip code."
4. Show effects are increasing in share of MW workers (we already do this)
   • "Showing that the estimated effect on local rents of a given increase in the minimum wage in the zip code of work increases linearly as a function of the share of workers in the zip code who are affected by changes in the minimum wage would increase the credibility of the paper."
5. Can you do something about consumption locations?
   • "residence minimum wage does not take into account where residents shop .... However, it would be useful to have a discussion of how this measurement error may bias the results."
   • "What share of consumption for the typical US household is from the zip code of residency? ... Can the authors at least separate zip codes of residence between those that have some retail in them and those that don't have any retail in them?"

Referee 3

1. Time horizon: change in housing demand is not so flexible.
   • "[Paper] assumes perfectly flexible adjustments along the intensive margin of housing consumption (i.e., floor space per capita)."
   • "If workers change their place of residence to adjust the floor space consumption, why do they have to stick to the original residential location?"
2. Over-emphasized theoretical contribution.
   • "Comes almost by assumption."
   • "If the authors would like to emphasize the theoretical novelty, I would appreciate discussions on how this model is different from Ahlfeldt et al. (2022)."
3. Error in statement of proposition 1?
   • "The proof shows that for any z′ with positive income shock, not for some z′. Moreover, the proof also shows that for z′ with no income shock, the rent effect is zero."
4. New contracts vs existing contracts.
   • "Assuming that posted prices are analogous to newly-contracted prices, this might explain why the rent seems to re- spond very fast to minimum wage changes in this paper."
5. Effect by housing type.
   • "In Section 5.4, significant rent effects are observed for types of units that are not occupied by poor workers. Why is this?"
6. Misinterpreting Agarwal et al (2022)?
   • "So the effect is much larger in Agarwal et al. (2022) than this paper. How to reconcile these two findings?"
7. Counterfactual.
   • Unclear what we gain with residence and workplace MW. "... the authors could compare the main result (Figure 6) with the spatial heterogeneity in the cases of considering workplace minimum wage only or residence minimum wage only."
   • "Suppose that local minimum wages are determined by homeowners in a decentralized way (c.f., Simon and Wilson 2021). ... It might be interesting to extend the model and consider a counterfactual scenario in which decentralized minimum wage rates are set in a Nash equilibrium."
8. Delineate the central districts in the city of Chicago, as not everyone will be familiar with it --> IMPLEMENT
9. "This paper aims to estimate two kind of treatment effects simultaneously. So it might suffer from the negative weights problem recently discussed in Goldsmith-Pinkham, Hull and Kolesar (2022)."

[santiagohermo]

Things we'll do from the reports:

1. Comments on plausibility of effects (R1.1, R3.1)
   • Clarify that our estimates refer to newly-contracted // posted rents, which are expected to response quickly. Existing contracts should respond more slowly.
2. Comments on selection (R1.2, R1.3).
   • Use Zillow listing count and Zillow rental index.
   • Search for alternative data sources. Can you show similar magnitude effects somewhere else?
     • HouseCanary; AHS. Promise to keep searching
3. Effect on prices (R1.4)
   • Unfortunately, data on consumption at the zip code level is not available. We can compare more with existing estimates.
4. Back of the envelope calculation using a guess on price effect (R1.5, R2.2)
   • We can definitely do this. Show that estimates are reasonable. Get estimates of consumption share from data.
   • Highlight more the share pocketed.
5. Plausibility of effect on income. (R1.6)
   • Highlight comparison to Cegniz et al 2019.
   • We could further show the robustness of these estimates if the editor thinks that would useful.
6. Household structure (R1.7, R3.5)
   • Acknowledge that models abstracts away from household structure.
   • Point to evidence that there exist many households with low-income earners.
   • Heterogeneity by household structure using census/ACS variables. Maybe income of household head.
7. Wage spillovers to contiguous jurisdictions (R2.3)
   • Don't the estimates already show an increase in income relative to contiguous jurisdictions? Unclear how to address. For now, ignore.
   • Use as outcome the wage bill relative to average in county
8. Show effect on income is increasing in share of MW workers (R2.4)
   • We will do this.
9. Can you do something about consumption locations?
   • We are not aware of estimates of the share of consumption in same ZIP code, although we will explore this more.
   • Heterogeneity based on importance of retail in zip code using ZIP code business patterns
10. Time horizon (R3.1)
    • Clarify that we don't need people not to change ZIP codes. We need the commuting shares to stay on average fixed.
    • Clarify that a bargaining model would imply similar results, and the mechanism wouldn't be an increasing a housing demand but an improvement in the outside option of renters.
11. Theoretical contribution. (R3.2, R3.3)
    • We are happy to point out more explicitly that the model is for motivation, and downsize our claims of a contribution.
    • We have a footnote (fn 16) where we argue why we think the model is a useful approximation even when homogeneity doesn't hold. We could make a new subsection discussing weaknesses of the model and the consequences of weakening the assumptions.
    • We can also argue that the residence MW is an approximation for a "consumption MW."
    • Compare more closely to Ahlfeldt et al. (2022).
    • Review proof proposition 1.
12. Effect by housing type. (R3.5).
    • It is unclear that those estimates imply an effect on houses not occupied by MW workers. Why? They are noisy, and we show that there are some MW workers in these categories. We can clarify these points.
13. Misinterpreting Agarwal et al. (2022)? (R3.6)
    • It is correct that we misread Figure 4. Our effects are smaller than Agarwal et al 2022. We will discuss in more detail the plausibility of our estimates based on the accounting exercise.
14. Problems with estimation (R3.8)
    • Make more emphasis on non-parametric exercise.
    • Ask professor Hull if there is a better way to address this.

[santiagohermo]

I'm done with a first version of the letter @diegogentilepassaro @gabrieleborg! It is here: cover.pdf

Can you take a read and make comments with Adobe Acrobat? I will also send the letter to professors for comments, which I will post here.

Once we collect comments and revise I will finish the submission! I'd be ok with meeting after that, but let me know if you want to meet before.

[santiagohermo]

Initial reactions

Neil

Looks great! A few initial comments on the response to the editor (I'll take a closer look at the rest later): - Point 1: I would suggest adding a clearer summary statement on top after summarizing the editor's concern. You can add something like, "We can address these concerns regarding the time dimension of the results by clarifying the timing of our rent measures and providing complementary evidence using alternative data sources." - Point 2: It sounds like you're saying that the editor had a concern about a robustness check, and they turned out to be correct. Upon a quick glance, this would seem to indicate a weakness in the paper. I think what you mean to convey is that the editor raised an issue regarding the interpretation of the results, and you agree with their interpretation, which actually makes for an interesting result that you will discuss in the paper. - Similar comment on point 3: Here it sounds like you will be qualifying the results and therefore weakening the paper. I would suggest taking this opportunity to state clearly to the editor what you view as the main contribution of the paper (presumably not providing direct evidence for that channel), and how the paper is still able to achieve its stated goals. Best, Neil

Jesse

Santiago, Thanks for sending this draft. I attach a version with some initial comments. A few broad themes: From the standpoint of the publication process, I see the goal of this step as to make the editor feel comfortable that a suitably revised version of the paper could eventually be publishable in this journal. For that reason, if there are important questions about the plausibility or sensitivity of your findings, then it is useful to provide enough information for the editor to evaluate their importance and understand concretely how you can address them. If someone has raised a question about the logic or accuracy of your assumptions, then it is important to either defend them affirmatively as reasonable, or change them. Arguing that you could in principle change the assumptions, or that they are not very important, is not enough. The editor will be looking for gaps or contradictions in your reasoning. So, if something you write seems not to address the referee's points, or to raise new questions or contradictions, that will not be helpful. If the previous editor or referees raise points that you think were already addressed in the previous draft, I would think again. Chances are that some new or clearer analysis is needed. Let me know what you think and we can discuss how best to proceed. Among the editors, all would be good I'm sure, but I might suggest Kory Kroft. -Jesse Attachment: [cover.jms.pdf](https://github.com/diegogentilepassaro/min_wage_rent/files/11149583/cover.jms.pdf) After replies and response to replies: [cover.jms.sh.jms.pdf](https://github.com/diegogentilepassaro/min_wage_rent/files/11153184/cover.jms.sh.jms.pdf)

[santiagohermo]

I'm making changes so that, in the letter, we move to make some arguments about standard errors. In this comment I paste some code to make a quick simulation and verify the claims in an example.

See code

```R remove(list=ls()) library(dplyr) library(fixest) dt <- data.frame(zip = rep(1:100, each = 12*5), year = rep(rep(2001:2005, each = 12), times = 100), month = rep(rep(1:12, times = 5), times = 100), date_id = rep(1:(12*5), times = 100)) dt <- dt %>% mutate(time_FE = date_id/100 - .5*(date_id/100)^2, zip_FE = zip/(100+.2*zip^2)) dt$x <- rnorm(n = nrow(dt)) dt$res <- rnorm(n = nrow(dt), sd = .25) dt <- dt %>% mutate(y = time_FE + zip_FE + .1*x + res) dt_yearly <- dt %>% group_by(zip, year) %>% summarise(y = mean(y), x = mean(x)) fit_monthly <- feols(y ~ x | zip + date_id, dt, cluster = ~zip) fit_yearly <- feols(y ~ x | zip + year, dt_yearly, cluster = ~zip) ````

[santiagohermo]

• Useful paper from 2020 that looks at the effect of the MW on consumption expenditure: https://onlinelibrary.wiley.com/doi/full/10.1111/jmcb.12684

• Lemos (2008) summarizes early literature estimating the effect of the MW on prices: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1467-6419.2007.00532.x

• Recent paper (2023) looking at the effect of the MW on poverty https://www.nber.org/papers/w31182

[santiagohermo]

I already got some comments from Jesse and Neil @diegogentilepassaro @gabrieleborg (see here), so I went ahead and implemented updates. The current version of the letter is cover.pdf

I won't make any changes till next week. If you have to take a look at this version, your comments would be great. Thanks!

[santiagohermo]

More reactions to updated letter:

Jesse

I already had some comments from Jesse on two points: - Argument about standard errors in yearly model - Argument about state by time FE (I'll open a new issue for this one) Newly annotated pdf: [cover.jms.pdf](https://github.com/diegogentilepassaro/min_wage_rent/files/11179305/cover.jms.pdf)

Peter H

Hi Santi See attached for some comments on your cover letter. I think your responses to the comments on the annual analysis and the state-by-time FE robustness check could be improved. The rest reads pretty sensibly to me Hope this is helpful, [cover_PDH.pdf](https://github.com/diegogentilepassaro/min_wage_rent/files/11190497/cover_PDH.pdf) [cover_PDH_SH.pdf](https://github.com/diegogentilepassaro/min_wage_rent/files/11236574/cover_PDH_SH.pdf)

Neil T

The revised discussion of the yearly model is excellent! I don't think there are any adjustments you need to make with regard to tone. While the new version is generally a clear improvement, the only part I found harder to follow is the updated point 2 in response to the editor (e.g., what is meant by "undesirable comparisons"). A small typo: "estimats"

[santiagohermo]

Do tenants bargain the posted price? The answer seems to be an overwhelming "no".

https://www.quora.com/How-different-is-the-rent-paid-by-a-tenant-and-the-rent-posted-online-of-the-same-housing-unit-Do-tenants-have-space-to-bargain-the-posted-price-or-is-it-common-for-tenants-to-just-accept-the-posted-price

[santiagohermo]

I got very detailed comments from JMS from the version in the latest commit (https://github.com/diegogentilepassaro/min_wage_rent/commit/bdc35265969a340cdb2cfcc242878fe46d091d02): SH.MW.JMS.pdf

I will update the letter based on this, so hopefully we'll submit very soon @diegogentilepassaro @gabrieleborg

[santiagohermo]

Paper to mention in the review by Cegniz et al 2022. It is useful to see the demographics characteristics of MW workers, the share of MW workers, etc.

[santiagohermo]

New version of the cover letter in https://github.com/diegogentilepassaro/min_wage_rent/commit/a9017ba76a0b380e7ccb2c7a9a114a6e07b7a69b. Drafftable diff with respect to previous version is here

[santiagohermo]

I'm submitting with the cover letter in https://github.com/diegogentilepassaro/min_wage_rent/commit/842e495cd41e227d7e0c0eb5948e2c4dd1fe5ffe today. Vamo arriba @diegogentilepassaro @gabrieleborg!

Note: The PDF proof of the submission looked good, but the HTML one is not working for some reason (computer goes on tilt when I try to download things). I'll just skip it.

Example


[santiagohermo]

Submission receipt: AEJPolicy_submission.pdf

[santiagohermo]

Summary: In this issue we discussed the AER rejection decision and re-submitted to the AEJ: Policy.

Submitted cover letter is from https://github.com/diegogentilepassaro/min_wage_rent/commit/842e495cd41e227d7e0c0eb5948e2c4dd1fe5ffe Submission receipt in https://github.com/diegogentilepassaro/min_wage_rent/issues/250#issuecomment-1525900049

Changes merged to master in https://github.com/diegogentilepassaro/min_wage_rent/commit/9e71e1db60affdad29ec6574aff5d23ca564f456