Abstract1

This paper studies how entrepreneurship in Colombia was affected by a nationwide policy that made almost half a million undocumented migrants from Venezuela eligible for resident’s visas (PEPProgram). The authors find that having a visa increased the levels of entrepreneurship among this population. As written in the evaluation [content] below, it has a lot of strengths. The contributions are clear and are both empirical (rarely studied outcome) and methodological (new empirical approach). In the “suggestions and minor suggestions” section, some possible extensions and [critiques] are specified.

Summary Measures

We asked evaluators to give some overall assessments, in addition to ratings across a range of criteria. See the evaluation summary “metrics” for a more detailed breakdown of this. See these ratings in the context of all Unjournal ratings, with some analysis, in our data presentation here.2

Overall assessment: We asked evaluators to rank this paper “heuristically” as a percentile “relative to all serious research in the same area that you have encountered in the last three years.” We requested they “consider all aspects of quality, credibility, importance to knowledge production, and importance to practice.”

Journal rank tier, normative rating (0-5): “On a ‘scale of journals’, what ‘quality of journal’ should this be published in? (See ranking tiers discussed here)” Note: 0= lowest/none, 5= highest/best”.

See here for the full evaluator guidelines, including further explanation of the requested ratings.

Written report3

First Impressions

This paper studies how entrepreneurship in Colombia was affected by a nationwide policy that made almost half a million undocumented migrants from Venezuela eligible for resident visas (PEP Program).

The authors find that having a visa increased the levels of entrepreneurship among this population.

Strengths

It is an amazing paper to read. Very well written, especially since the context of the policy is key for understanding the logic of the methodology used.

Its contributions are both empirical and methodological. In the empirical part, it studies entrepreneurship, an outcome that has not been studied much in the literature of gains from legalizing migration. The methodological contribution is the application of the new methods that study how to get casual estimates when there is a nonrandom exposure to an exogenous shock. They translate these methods to the context of a regression discontinuity design and propose an instrument for differences between the predicted and actual realizations of the running variable. This is one of the most novel parts of the article.

Apart from the proposed methodology and its instrument, it also uses a regular panel method which provides similar results for the main specification.

Suggestions

I find it hard to make comments criticizing this paper, however, some ideas have come to mind that might be useful to think about them, either as a robustness check or a possible extension.

1. The article finds “similar impacts on the creation of both employer and non-employer firms.” While these employer firms create 1 to 6 new jobs and are not ‘high growth’ by the typical standards of developed countries, they still represent meaningful economic spillovers.” I wonder if it would be possible to try to study some of these economic spillovers. For example, using employer-employee matched data from the “Planilla Integrada de Liquidación de Aportes (PILA)” one could study the effect of obtaining the PEP Pardon on the likelihood of obtaining a job in these new firms created by Venezuelans. If there is an effect, it would be related to the literature on the importance of networks in the job market like for example Calvó-Armengol (2004)[1] or Kramarz and Skans (2014)[2].

2. In the panel specification, basically it’s like a diff in diff approach between those who received and those who did not receive PEP. It is not that clear to me why those who did not receive it see an increase (albeit much lower) in entrepreneurship (see figure 7). Did something change?

3. If there is also the date on where these persons received the visa, one could also use the panel data for an event study instead of a diff in diff, as a robustness check.

4. The treatment effect by year seems to be statistically significant until 2022 and 2021 (figure 6). I was wondering about the mechanism that might explain this lagged effect.4 My prior is that to start a business, although having access to the banking and borrowing systems are important, having a credit history in a bank is as important and takes longer. The effects of the event study might be comparable with the findings on figure 6. However, this might also be because of the Covid.

5. This is speculative, but you can use the identification of Bahar, Ibañez and Rozo (2021)[3] at the municipal level to see aggregate effects on entrepreneurship. If it’s only the sum of firms of persons with the PEP pardon, it would be a robustness check in the aggregate of some sort. If it is new firms in general, it would include the direct effect of having more entrepreneurs and the indirect effect of having a larger pool of workers, which might benefit all firms because of potentially cheaper labor. Although, given figure 6, it seems there is no effect for Colombian native firms.

6. An improvement in the identification of papers like Bahar, Ibañez and Rozo (2021)[3] might be done also. In there, they use as an instrument the average registration day available by department. One could compute not only the actual average registration days available by department but the expected one, using the same permutations done in the paper.

Minor Suggestions

1. In the introduction, in the first page, the authors argue that in the panel data approach, they can decompose the immigrant entrepreneurship effect into a physical relocation effect and a visa effect. I firstly understood it as if the PEP-Pardon had an effect in migration towards different cities (for example, if I am going to open a business, maybe a big city like Medellín or Bogotá would be better). Then later in the text they explain that the physical reallocation effect is moving from Venezuela to Colombia. It could be my reading, but it might be worth it to recheck it.

2. Although it is a very good idea to present the assumptions in a general context… It would [also] be useful if at the end of each assumption, you could state for the case of the paper what would that mean directly. It would help the reader to think straightforwardly if the assumption is getting realized in the case of the study and to think about potential threats in a more direct way.

3. Table 8 and Table 9 show results for different sub-samples. Here, the results in the panel regression differs from the results using the IV cross-section. It would be important to explain these differences.

Main Advice

Regarding the identification, the regression discontinuity design compares outcomes among migrants who completed the RAMV, some earlier than other, obtaining a greater time window for the legalization process. Now, since both completing the RAMV earlier and having more time to obtain the visa might be endogenous to the level of entrepreneurship of the person, a source of randomness like precipitation is cleverly used. In terms of the assumptions for the instrument to work, they are well explained, and the context is a context where the instrument can work. However, the strategy is based on compliers whose PEP status was sensitive to the timing and length of the window.

This is the part of the paper that might be harder to believe. Even the people with a smaller time window had more than two months to get the PEP visa. One could think having two months or having two months and one week would not make such a difference, given the perks of being legalized in a country. Although table 3 tries to compare a classification of always takers vs compliers, one thing that could help is if the document could talk about the time people took to obtain the PEP visa. By this I mean the number of days from the date they could start the process, to the date they submitted the documents online to get the PEP visa. This would help to get an idea of how important is to have one or two weeks more, for example.

Conclusions

Overall, I find the paper very interesting, both methodologically and empirically. It is very well done, and I would think it is very likely to be published in a very good journal. I am not an expert in this field nor in the econometric procedure, but I would start by aiming, with a paper like this, for a top general interest journal and see the comments and feedback from those type of reviewers.

References

[1]Calvó-Armengol, Antoni. 2004. “Job contact networks.” Journal of economic Theory, 115(1): 191– 206.

[2]Kramarz, Francis, and Oskar Nordström Skans. 2014. “When strong ties are strong: Networks and youth labour market entry.” Review of Economic Studies, 81(3): 1164–1200.

[3]Bahar, Dany, Ana María Ibàñez, and Sandra V Rozo. 2021. “Give me your tired and your poor: Impact of a large-scale amnesty program for undocumented refugees.” Journal of Development Economics, 151: 102652.

Evaluator details

1. How long have you been in this field?

   • My research focuses on applied work but not particularly related to labor or migration. So I would say, not so long ago.

2. How many proposals and papers have you evaluated?

   • 3 journals, 9 master thesis.