Political Entrepreneurs and Pork-Barrel Spending

Round 1

Reviewer 1 Report

Where the Intro moves to political entrepreneurs (line 47) in paragraph four, the Intro section changes direction and morphs into something like a literature review.  The author(s) should clear this up by creating a new section 2 that discusses the literature.

The paper seems to ignore the conceptual work on rent seeking by Tullock and others (Krueger, etc.), as well as the many empirical studies investigating Tullock’s rent seeking hypothesis (e.g., Sobel, Laband, etc.).  These seem central to the author(s)’ work here and some space should be devoted to them, no?

Author(s) need to better explain on page 3 why benefits to a state are likely to be more concentrated than benefits to a Congressional district when the opposite seems true.

The author(s) need to formally specify the proposed regression equation somewhere in the methodology section, and then go through it, variable by variable, stating the testable hypotheses.

The discussion about the federal aid to state and local governments brings to mind a paper by Hobson in Contemporary Economic Policy sometime around 2000 that should likely receive a mention here.

Perhaps the author(s) should integrate the idea of “name recognition” into their explanation of the “second surprise” result.

The author(s) should consider an additional variable or two in order to further explore the pork barrel electoral cycle that is discussed.  Would some interaction terms help?

The whole paper could be much better if the author(s) re-write much of it with attention to grammar (e.g., not ending sentences with prepositions, etc.), presentation (improve overall flow), and crispness of the discussion.  Please spend some significant time doing this as the quality of the piece depends on it.

Lastly, the author(s) should be clearer on how this paper fits in the literature.  What is done here that is not done before?  What are the new and unique findings?

Author Response

(Note: The same comments are included in the attached Word document, but it did not seem to be uploading correctly.)

Political Entrepreneurs and Pork-Barrel Spending

Response to Reviewer #1

Thank you for the critique of my paper. I have made rather major revisions which I will detail below. While there are a few other things that could help on the margin, I accomplished as much as I could do with the ten day turnaround requested by the editor. Overall, I made two major changes. The first is an extension of the data set. The original draft went through the 2012 election, but I extended the study through the 2018 election. While the level of pork-barrel spending has fallen, the added data points seem to help with the explanatory variables. The second is a rework of the prose (especially in the introduction of the paper). From the suggestions I received from both reviewers, I still tell a story of incumbents and political entrepreneurs, but I use rent-seeking and Mayhew's tools (credit claiming, advertising, and position taking.)

Below is a point-by-point reply. Your suggestions, comments, and concerns are in bold with my responses below.

Where the Intro moves to political entrepreneurs (line 47) in paragraph four, the Intro section changes direction and morphs into something like a literature review.  The author(s) should clear this up by creating a new section 2 that discusses the literature.

Thank you for the suggestion. I have reworked section 1 by splitting it into three sections. This should help add order and flow to the work. According to the editor, section 2 is reserved for data, so I added my literature review to section 1 instead.

The paper seems to ignore the conceptual work on rent seeking by Tullock and others (Krueger, etc.), as well as the many empirical studies investigating Tullock’s rent seeking hypothesis (e.g., Sobel, Laband, etc.).  These seem central to the author(s)’ work here and some space should be devoted to them, no?

I have reworked the prose to incorporate this idea that should have originally been included and, in my opinion, adds a good deal to the story I am trying to tell. The changes are especially prominent in sections 1 and 2. In fact, a subsection dealing with rent-seeking has been added to section 2.

Author(s) need to better explain on page 3 why benefits to a state are likely to be more concentrated than benefits to a Congressional district when the opposite seems true.

My original wording was very confusing and did not accurately describe what I meant. I have reworded that section. The changes can be found in lines 146-151. My point was more about the capture of benefits. Specifically, a project within a district will typically have its benefits spill over to other districts as well whereas a state is able to capture a larger percentage due to being a larger geographical area. If I were doing a spatial study I would dedicate more to this (since projects completed near a state line will also leak benefits to other states). But I hope my original point has been better explained. 

The author(s) need to formally specify the proposed regression equation somewhere in the methodology section, and then go through it, variable by variable, stating the testable hypotheses.

This is now completed. I laid out the base form of the regression (since each regression follows the same structure). In addition, I have added several subsections to section 2 where I go through each variable and discuss the hypothesized relationship. I feel like this adds some crispness to the paper as I have broken up a long-winded section into pieces that are a bit more organized and easier to digest.

The discussion about the federal aid to state and local governments brings to mind a paper by Hobson in Contemporary Economic Policy sometime around 2000 that should likely receive a mention here.

Thank you for the suggestion. This has been incorporated in lines 273-277.

Perhaps the author(s) should integrate the idea of “name recognition” into their explanation of the “second surprise” result.

Again, thank you for the idea. I have included discussion on this (and a general discussion on seat vulnerability) in the discussion of several variables including in lines 312-317.

The author(s) should consider an additional variable or two in order to further explore the pork barrel electoral cycle that is discussed.  Would some interaction terms help?

This is, unfortunately, one change I was unable to make. With the ten days given by the editor to turn the paper around, I only had time to extend my data set up to and including the most recent 2018 Senate election. I did attempt to use the existing variables to look for interaction. The only relationship with any promise (though still not close to be significant) was pork variable and the “same party” variable (the binary that has a value of one if the Senators belong to the same party.) If the interaction term had worked, it would have shown that Senators can amplify their campaign contributions by using pork-barrel spending when done in cooperation with a Senator from the same party. This could have also added to the purple state idea presented in this draft of the paper. I also tried to interact several of the other variables, mainly with the pork variables, but was unable to uncover any new relationships. I am sure there are other useful variables out there that could add to the study, but I was unable to find anything in the 10 days given.

The whole paper could be much better if the author(s) re-write much of it with attention to grammar (e.g., not ending sentences with prepositions, etc.), presentation (improve overall flow), and crispness of the discussion.  Please spend some significant time doing this as the quality of the piece depends on it.

I spent a considerable amount of time on this. I hope that the re-organization of sections 1 and 2 help with the flow and crispness of the paper. I did have “track changes” enabled, so you can see the extent of the revisions. There may be a paragraph or two that went untouched, but the vast majority of the paper was rewritten.

 Lastly, the author(s) should be clearer on how this paper fits in the literature.  What is done here that is not done before?  What are the new and unique findings?

I included discussion of my contribution in both the introduction (lines 38-47) and conclusion (lines 494-498). I hope that this gives a clearer idea of my contribution to the existing literature.

Again, thank you for all of your suggestions. I truly believe that this paper has been greatly improved thanks to all of your feedback!

Author Response File: Author Response.docx

Reviewer 2 Report

Political Entrepreneurs and Pork-Barrel Spending Review

This article presents an interesting argument.  However, I have some serious concerns about the theoretical grounding of the work as well as some missing variables.  I will detail my concerns below.  With substantial revisions the manuscript could be suitable for publication in Economies.

First, there are some foundational works in Political Science upon which the author draws that are in no way cited in the paper.  David Mayhew’s essential work Congress: The Electoral Connection argues that members participate in three main activities to insure their reelection: Advertising, Credit-Claiming and Position Taking.  This manuscript draws heavily on the concept of credit-claiming, yet it is not mentioned anywhere, and Mayhew is not cited.  Similarly, to discuss campaign finance spending and never reference the many works of Gary Jacobson on the topic is concerning as well. 

In terms of the analysis, I have concerns about the explanations of variables and possible bias from missing variables.  Near line 188 there is discussion of the margin of victory in past elections.  I think the author needs to more explicitly state that vulnerability affects the need to fundraise.  While this idea seems obvious the author is never clear that past electoral margin is included because it likely has a direct effect on fundraising.  Similarly, on line 269 where the author says its surprising that past success yields lower fundraising totals, I do not find this surprising at all.  A senator who routinely wins reelection by large margins has no need to raise large sums of money because they are in a safe seat. 

Also, on page five the author claims that senators from larger states are more likely to raise more money because they have a larger in-state donor base.  I disagree. Senators from larger states raise more money because its more expensive to run in those large states.  Similarly, competitive races in small states raise significant amounts of money because money flows into the state from out-of-state donors. 

The justification for the gender variable also feels incredibly dated.  I would like to see a more current justification for including gender as a variable predicting fundraising.  Most current research suggests that when women run they win at the same rates are not disadvantaged. 

Finally, I’d just like to see more discussion of the fact that there are two senators who could both benefit from securing pork barrel spending.  How is this accounted for?  How does it play out in states with one republican and one democratic senator?  Is the effect the same in these purple states?

In terms of the conclusion, the author needs to be careful to not assert causation.  The findings here have not proved a direct link only a possible relationship.  It might also be helpful to return to the idea of political entrepreneurs in the conclusions.  This idea is introduced early on but then never revisited. 

Finally a few smaller notes, in the methodology, around line 117 in the manuscript the author writes: “The results of this paper suggest that Stratman has misidentified the relationship and that Samuels analysis in more descriptive”.  This sentence is confusing at this point in the paper.  Are you referencing the main findings of your work?

One final note is that I’d remove discussion of presidential fundraising as its very difficult to compare fundraising for the presidency to senate fundraising. 

I also have significant concerns about the author relying heavily on findings from Brazil as the country has a very different political climate. 

A brief discussion of congressional attempts to limit pork barrel spending would also be helpful. 

Author Response

(Note: The same comments are included in the attached Word document, but it did not seem to be uploading correctly.)

Political Entrepreneurs and Pork-Barrel Spending

Response to Reviewer #2

Thank you for the critique of my paper. I have made rather major revisions which I will detail below. While there are a few other things that could help on the margin, I accomplished as much as I could do with the ten day turnaround requested by the editor. Overall, I made two major changes. The first is an extension of the data set. The original draft went through the 2012 election, but I extended the study through the 2018 election. While the level of pork-barrel spending has fallen, the added data points seem to help with the explanatory variables. The second is a rework of the prose (especially in the introduction of the paper). From the suggestions I received from both reviewers, I still tell a story of incumbents and political entrepreneurs, but I use rent-seeking and Mayhew's tools (credit claiming, advertising, and position taking.) I spent a considerable rewriting the paper as well. I hope that the re-organization of sections 1 and 2 help with the flow and crispness of the paper. I did have “track changes” enabled, so you can see the extent of the revisions. There may be a paragraph or two that went untouched, but the vast majority of the paper was rewritten.

Below is a point-by-point reply. Your suggestions, comments, and concerns are in bold with my responses below.

This article presents an interesting argument.  However, I have some serious concerns about the theoretical grounding of the work as well as some missing variables.  I will detail my concerns below.  With substantial revisions the manuscript could be suitable for publication in Economies.

First, there are some foundational works in Political Science upon which the author draws that are in no way cited in the paper.  David Mayhew’s essential work Congress: The Electoral Connection argues that members participate in three main activities to insure their reelection: Advertising, Credit-Claiming and Position Taking.  This manuscript draws heavily on the concept of credit-claiming, yet it is not mentioned anywhere, and Mayhew is not cited.  Similarly, to discuss campaign finance spending and never reference the many works of Gary Jacobson on the topic is concerning as well.

Thank you for the concern as I believe it (and a similar concern about the inclusion of rent-seeking literature from the other reviewer) helped me restructure the prose. It is my hope that the inclusion of all three pools of literature improve the overall substance, style, and theoretical underpinnings. The inclusion of Mayhew now begins on line 1 and is seen throughout. Jacobson is mentioned in lines 84-90.

In terms of the analysis, I have concerns about the explanations of variables and possible bias from missing variables.  Near line 188 there is discussion of the margin of victory in past elections.  I think the author needs to more explicitly state that vulnerability affects the need to fundraise.  While this idea seems obvious the author is never clear that past electoral margin is included because it likely has a direct effect on fundraising.  Similarly, on line 269 where the author says its surprising that past success yields lower fundraising totals, I do not find this surprising at all.  A senator who routinely wins reelection by large margins has no need to raise large sums of money because they are in a safe seat.

I have changed the way that I approach this idea. In section 2.4 (explanatory variables), I discuss how vulnerability impacts fundraising. This comes through in three variables. Vulnerability is discussed under challenger fundraising (lines 290-294), the performance in previous elections variable (lines 301-309), and terms served (lines 312-317). I hope that this idea comes across far more clear than it did previously. I agree completely with your sentiment that it was alluded to but never actual stated. I included the same logic in a few other places as well in the results as well.

Also, on page five the author claims that senators from larger states are more likely to raise more money because they have a larger in-state donor base.  I disagree. Senators from larger states raise more money because its more expensive to run in those large states.  Similarly, competitive races in small states raise significant amounts of money because money flows into the state from out-of-state donors.

Because I am now only using a fixed-effect model, I was able to exclude population and was able to eliminate this discussion. I do agree with your sentiment that larger states are more expensive (and shocked by the fact that nearly $125 million was recently spent on the Texas Senate race between Cruz and O’Rourke!)

The justification for the gender variable also feels incredibly dated.  I would like to see a more current justification for including gender as a variable predicting fundraising.  Most current research suggests that when women run they win at the same rates are not disadvantaged.

This has also been changed. It came across as laziness on my part. I have seen the female variable used in other papers so I included it. When I went to cite a paper to defend its usage, I obviously picked a dated article that doesn’t apply anymore. While the original citation is still included to show historical precedent, the newer work I found says that female candidates do fare about equally as well as males, but that there is still bias that needs to be overcome (even if it is unintentional bias). This actually comes across in my results as being female leads to additional fundraising, but I believe this is because a female may need to do a bit more to overcome the bias. Successful female candidates accomplish this and compete well against male candidates.

Finally, I’d just like to see more discussion of the fact that there are two senators who could both benefit from securing pork barrel spending.  How is this accounted for?  How does it play out in states with one republican and one democratic senator?  Is the effect the same in these purple states?

I have expanded the discussion of this throughout the paper. I do talk about the flaw with a state-level analysis in lines 152-158 and lines 226-231. The attempt to explore the relationship between the two is done through the same party variable. This variable is discussed in lines 328-334. While it doesn’t appear that belonging to the same party brings in additional campaign contributions, it does appear that it does increase the percent of total fundraising attributed to the incumbent. This likely speaks again to the idea of seat vulnerability. Incumbents are far more vulnerable in purple states compared to red or blue states. In fact, the largest source of vulnerability is primary elections which are not included in this study.

In addition, per a suggestion from the other reviewer, I tried to include interaction terms including one between pork-barrel spending and whether the Senators belonged to the same party, but nothing of significance came to be.

As I mentioned, the fact that this is a state-level analysis does have its weaknesses, but it is the only way to explore the impact of pork on campaign fundraising over such a long period of time.

In terms of the conclusion, the author needs to be careful to not assert causation.  The findings here have not proved a direct link only a possible relationship.  It might also be helpful to return to the idea of political entrepreneurs in the conclusions.  This idea is introduced early on but then never revisited.

I completely agree. I can’t count the number of times I have taught this concept and I definitely crossed the line. I changed the way I interpret my results in the conclusion and also tie together the earlier discussion on Mayhew, rent-seeking, and political entrepreneurs.

Finally a few smaller notes, in the methodology, around line 117 in the manuscript the author writes: “The results of this paper suggest that Stratman has misidentified the relationship and that Samuels analysis in more descriptive”.  This sentence is confusing at this point in the paper.  Are you referencing the main findings of your work?

I have revised this discussion (now around lines 139-141). I attempted to tie in that idea to the rest of my argument.

One final note is that I’d remove discussion of presidential fundraising as its very difficult to compare fundraising for the presidency to senate fundraising.

This has been removed. I have added a few other examples in order to motivate at times, but they are all Senate elections (such as Robert C. Byrd and the Cruz/O’Rourke election.)

I also have significant concerns about the author relying heavily on findings from Brazil as the country has a very different political climate.

I have rewritten the sections where Brazil is discussed and I make it clear that while the political systems are different (especially in regards to the way kickbacks are treated), the goal of the paper is to explore the motivations of the incumbents which are the same in any (democratic) political system. This discussion is had in lines 105-108 and lines 174-178. While I use the idea in the Samuels paper, the variables are different and more in line with past research on American elections.

A brief discussion of congressional attempts to limit pork barrel spending would also be helpful.

This is now discussed in a couple of spots (lines 169-173 and lines 237-239). In addition, I have included two figures (around line 240) that shows the overall reduction in pork-barrel spending due to the 2011 moratorium.

Again, thank you for all of your suggestions. I truly believe that this paper has been greatly improved thanks to all of your feedback!

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The author has done a fine job responding to my earlier criticisms.  This is now a nice contribution to the special issue.

Reviewer 2 Report

I appreciate the author's efforts undertaken to revise the manuscript in a limited amount of time.  I am pleased with how each of my concerns has been addressed.  I appreciate the thorough revision that has clearly been undertaken and the detailed response to my concerns.  

I completely understand the use of track changes to identify the addition of new passages, however it makes proofreading for typos more difficult.  I would recommend that the author and editor undertake one final careful proofreading of the paper.  I am happy to recommend it for publication.