American Journal Of Health Economics

Why Your Stata Code Is Failing Replication Checks at AJHE (And What To Fix First)

The American Journal Of Health Economics follows a standard double-blind peer review process like most economics journals, but their replication requirements are stricter than you will expect if you have only submitted to applied health policy outlets before. When I submitted a paper on hospital price transparency effects last year, my initial reviewer comments came back asking for a fully executable do-file, not just the summary statistics from the main table. The problem was not that my results were wrong, but that my code used a local file path and a private SAS dataset that could not be shared under IRB restrictions. I had to restructure the entire workflow around publicly available Medicare claims data and write a completely separate version of the analysis script that produced identical coefficient estimates using only open sources. This type of situation is common enough that every economics PhD student should understand how the replication pipeline actually works before they bother drafting a cover letter.

American Journal Of Health Economics: Where It Fits and Who Actually Reads It

The Southern Economic Association publishes this journal, and it sits somewhere between purely theoretical health economics papers and the more policy-heavy Applied Health Economics and Health Policy journal. The distinction matters because your submission needs to signal that it contributes something to the economics side of the field, not just the medical side. Reviewers will look for clear identification strategies, proper handling of endogeneity, and robustness checks that go beyond throwing three alternative specifications at the wall. The most underrated insight about getting published here is that the editorial desk will reject a paper for methodology flaws faster than they will send it out for review. I watched a colleague submit a paper that used an instrumental variable approach with a weak first stage, and the editor did not even send it out. They wrote a two-paragraph rejection letter pointing directly to the F-statistic and suggesting the author cite Angrist and Pischke on identification. That is the kind of gatekeeping that saves everyone time, but it also means you should run your own diagnostics before anything reaches the editorial office.

How to Find and Access Articles Efficiently

Most people find this journal through Wiley Online Library since that is the current publisher hosting it, but the pricing model is a mess if you do not have institutional access. A single article subscription runs roughly forty dollars, which adds up fast when you are doing literature reviews for a dissertation chapter. If you are a graduate student, your university library subscription should cover this already, and you can usually download PDFs directly through your institution proxy without running into paywalls. If you are an independent researcher without institutional access, the options are limited to interlibrary loan requests, which typically take five to seven business days, or purchasing individual articles. The journal itself does not offer a formal open access option for subscriptions unless your institution participates in a read-and-publish agreement through Wiley. Some universities have negotiated this, so checking with your library is worth fifteen minutes of your time before you pay out of pocket. I learned this the hard way when I needed three specific articles for a conference presentation and ended up spending about two hundred dollars on direct downloads from the publisher site. The library database search returned the metadata but the full text links were blocked for anyone without an active subscription token.

What Makes a Submission Actually Survive Review

The biggest mistake I see is authors who treat the data and methods section as an afterthought. In health economics, this is where papers fail. Reviewers will scrutinize your sample construction, your inclusion and exclusion criteria, and exactly how you handled missing variables in claim-based datasets. If you drop observations without reporting how many were lost at each step, you will get hammered in review. I once had a reviewer ask me to report the exact number of observations dropped at each stage of my Medicare claims linkage, and I had to go back and reconstruct the full attrition table because my original draft just said the final sample consisted of approximately twelve thousand patients. Another thing nobody warns you about is the difference between statistical significance and clinical relevance in this journal. You can have a perfectly identified regression with a tiny standard error and a p-value of zero point zero zero one, and the reviewers will still question whether the magnitude of your coefficient matters in practice. A reduction of point zero zero zero three in readmission rates might be statistically significant with enough observations, but it is clinically irrelevant. State your findings in real world terms early in the paper so reviewers do not have to do it for you.

The Practical Downsides You Should Know About

The peer review timeline for this journal is reasonable but not fast. Expect four to six months from submission to first decision if your paper survives the desk rejection screen. The revision process can add another two to three months depending on how many reviewer rounds you go through. If you are working toward a graduation deadline or a tenure clock, this is important to factor into your planning. I knew someone who submitted a paper right before qualifying exams and waited eight months for a final accept, which left a gap in their publication record that looked bad on the job market despite the eventual outcome being positive. The journal also has a relatively high rejection rate at the desk level, which is frustrating but understandable given the volume of submissions they receive. Papers that rely entirely on descriptive analysis without any causal identification framework tend to get desk rejected quickly. If your contribution is primarily generating new descriptive statistics about a population, consider targeting a different journal that specializes in health services research rather than health economics. The Health Affairs journal or Medical Care might be better fits for purely descriptive work, even though the impact factors are comparable.

What the Replication Package Actually Needs to Include

AJHE requires a replication package that contains everything a competent econometrician would need to reproduce your main results. This means the data dictionary, the analysis code, and documentation that explains any non-obvious steps in your cleaning process. If you used any proprietary databases that cannot be shared publicly, you need to provide a detailed memo explaining exactly what data you used, where you got it, and what restrictions you had, so reviewers can assess whether your findings would hold with alternative datasets. This is not optional. Papers that send incomplete replication packages often get reviews that note the missing pieces as a serious concern, which slows the entire process down. One specific edge case that caught me off guard was my use of restricted IRS tax records for income variables in a health expenditure model. The tax data could not be shared at all, even in de-identified form, because of the confidentiality agreements attached to those records. I ended up writing a supplemental technical appendix that described the variable construction in enough detail for reviewers to evaluate whether my proxy variables were adequate, and I provided a sensitivity analysis showing that substituting census-level income estimates produced qualitatively similar results. The reviewers accepted this workaround, but it cost me an extra month of revisions and three additional drafts of the appendix. Understanding the mechanics of this journal takes more than reading the author guidelines. It comes from watching papers get rejected for reasons that are not spelled out anywhere in the submission instructions and learning to anticipate those problems before they happen. The field rewards methodological rigor and honest reporting of limitations far more than it rewards ambitious claims built on thin evidence.