Authorship disputes in international collaborations rarely come from bad intentions. They come from an asymmetry: one side has more institutional weight, and the conversation happens too late for that to be irrelevant.
Agree criteria at planning stage
Before data collection, not after. At that point nobody knows what the results will be, so both sides are negotiating principles rather than a known prize.
You do not need a name order. You need: what level of contribution earns authorship; who leads which type of output; who is corresponding author; and who arbitrates disagreement.
Put it in a one-page memorandum. Verbal understandings diverge over eighteen months, and they diverge in favour of whoever is remembering.
As the party with more institutional weight, you should be the one raising this. An early-career partner asking a senior foreign collaborator about authorship is taking a risk; you are not.
Three patterns to check yourself against
The partner collects all the data and lands in the middle of the author list. If your collaborator's team generated the dataset, the fix is not a better name position — it is involving them in design and in writing, so their position is earned under the criteria rather than negotiated as a courtesy.
A second paper from the same data without them. Agree explicitly which uses beyond the joint output require whose consent. This is the most common source of lasting resentment.
Substantive contribution acknowledged rather than authored. Read your own acknowledgements section and ask whether anyone in it did work that meets the authorship criteria. If so, that is a paper you should fix before submission.
A note on local practice
You may encounter requests to add someone senior who did not contribute, or to include a departmental head. This conflicts with standard authorship criteria and with most journals' policies.
Handle it by pointing to the journal's published criteria rather than by criticising local practice, and have the conversation with your collaborator privately. Putting them in a position of refusing their own institution's expectation, in writing, in front of others, is not a favour.
Equally, do not use this as grounds to dismiss the contributions of people who did do the work.
Corresponding author
This role is not administrative. The corresponding author holds the data, fields post-publication queries, and in practice controls the paper.
Defaulting it to the foreign partner on every output is a choice worth examining. If the dataset was generated in Vietnam and your partner's group holds the originals, having them lead at least one paper as corresponding author is both reasonable and a meaningful signal about the partnership.
It also has a practical benefit: it builds the partner's track record, which is what makes them a stronger collaborator on the next grant.
Contribution statements
Where the journal requires a detailed statement of who did what, fill it in accurately and have every author check their own entry.
These statements help partners whose contribution is fieldwork and data generation, because they record the actual work rather than leaving name order to carry all the meaning. Treat completing them properly as part of the writing, not as a form.
Data rights beyond the paper
Agree, in writing, before collection: who holds the originals, who may publish from them afterwards, for how long, and whether the data will be made open.
If your funder mandates open data, that must be settled with your partner and reflected in participant consent before collection begins. Discovering it at submission puts your partner in an impossible position.
Continuing access to the dataset is often the single most valuable thing the collaboration can leave with a partner group — it is what their next students work on. Treating it as a normal part of the agreement rather than a concession is the practical test of whether the partnership is equitable.
If a dispute arises
Raise it directly and cite the agreement and the journal's criteria. Keep it about the work.
If the paper has not been submitted, pause submission until it is resolved — proceeding while a co-author objects creates a much larger problem later.
If it is already submitted or published, journals and research institutions have formal processes. Engage them with documents: drafts, correspondence, the agreement.
And if you find yourself in a dispute where you hold the greater institutional power, be aware that "we could not reach agreement" is not a neutral outcome — it usually means the other party gave up.
When should authorship be agreed?
At planning stage, before data collection. Both sides are then negotiating principles rather than a specific paper whose value is already visible.
Who should raise the authorship conversation?
The party with more institutional weight. An early-career partner raising it with a senior foreign collaborator is taking a risk that you are not.
What if we are asked to add someone who did not contribute?
Point to the journal's published authorship criteria, and have the conversation privately with your collaborator rather than in a group setting.
Should the corresponding author always be the foreign partner?
No, and defaulting to it on every output is worth examining. Having the partner lead at least one paper builds the track record that makes them stronger on the next grant.
