Comments on abstract, title, references

The goal of the study and the findings are clearly stated in the abstract. The research question is very interesting and novel. It is assumed that the paper incorporates both binary and continuous evaluations of peacekeeping missions. While the title is informative, it is not fully relevant to the research methodology used in the study as the phrase `tangled-up’ only refers to the home-country composition of the peacekeepers, leaving the binary evaluation and the size of peacekeepers out. Besides that, the reference part seems correct. The references are relevant, recent and correctly cited.

Comments on introduction/background

The authors introduce the topic and research question clearly. The previous knowledge on the mechanism is supported with relevant citations.

The authors bring an alternative explanation to post-civil recovery and therefore the paper has a feature of theoretical novelty. The main argument in the paper is that the UN peacekeeping decrease the political violence through encouraging nonviolent mobilizations as the civil society internalize democratic norms easily when peacekeepers are present, which is referred as bottom-up mechanism.

The importance of the paper is well-outlined. However, the authors solely talk about paper's potential contributions to the existing literature. That would add up to the popularity of the paper more if the authors talked about the real-world benefits of this study as well.

Another thing which is not clear in the introduction is that the authors challenge the top-down explanations of the relation between peacekeepers and civil war recovery, and bring a novel explanation which is bottom-up. In this respect, I keep wondering whether they will be controlling for previously accepted top-down factors or whether their findings for the bottom-up factors will be majorly driven by the top-down factors. The empirical strategy adopted for this specific issue should have been better explained in the literature.

Comments on methodology

The authors did a good job in explaining the process of case selection. For the four hypotheses of the paper, the authors adopt four different empirical strategies which are valid and reliable. The dependent variable, four independent variables and control variables are all well defined and measured. The graphs and descriptive statistics exist to show the distribution and range of the variables. The study is fully replicable.

The fourth hypothesis is the most important contribution of the paper to the literature. The authors emphasize that point one more time in the methodology part that I believe is a good way of promoting the importance of the study.

While I find research design and data analysis sections very promising, I have some comments that I believe might improve the understanding of the text.

1. The X axis of Figure 2 is Civil Society Score. Y axis of Figure 3 is also Civil Society Score. These variables are not the same. The former represents the civil-society scores for TCCs, while the latter represents civil-society PK scores of peacekeeping missions. In other words, one is a composition variable and the other is the variable used in the calculation of the composition variable. Thus, the x and y axis should be renamed in order to avoid confusion.

2. To better visualize the distribution of the civil society PK scores, the simple histogram in Figure 3 may be replaced by a xy plot showing the scores of each PK mission at different time points.

3. The last paragraph of control variables ends with the sentence that "Table 1 below contains descriptive statistics...". However, in the text quantitative analysis comes immediately after this section and Table 1 seems that it has floated to the next session. Therefore, Table 1 should be repositioned.

4. It is unnecessary that the authors conduct correlation check in Model 1, 3, 5 and 8. These specifications should be in the Appendix and there is no point in showing them in the main tables.

5. It is not well-explained in the text what the subsetted sample is composed of in Model 7 and 10. Thus, I cannot comment on the empirical findings in those models.

The conclusion section satisfies all of the expectations.

1. The results are discussed avoiding a repetitive language.
2. Policy recommendations and notes for future research are well outlined.