Are the Contents of International Treaties Copied and Pasted?

Most accounts of international negotiations suggest that global agreements are individually crafted and distinct, while some emerging scholarship suggests a heavy reliance on models and templates. In this research, we present a comprehensive test of whether new international treaties are heavily copied and pasted from past ones. We specify several reasons to expect widespread copying and pasting, and argue that both the most and least powerful countries should be most likely to do so. Using text analysis to examine several hundred preferential trade agreements (PTAs), we reveal that most PTAs copy a sizable majority of their content word for word from an earlier agreement. At least one hundred PTAs take 80 percent or more of their contents directly from a single, existing treaty—with many copying and pasting 95 percent or more. These numbers climb even higher when we compare important substantive chapters of trade agreements, many of which are copied and pasted verbatim. Such copying and pasting is most prevalent among low-capacity governments that lean heavily on existing templates, and powerful states that desire to spread their preferred rules globally. This widespread replication of existing treaty language reshapes how we think about international cooperation, and it has important implications for literatures on institutional design, policy diffusion, state power, and legal fragmentation.

That is from a new paper by Todd Allee and Manfred Elsig, via the excellent Kevin Lewis.


I've long felt that writing treaties and contracts is much like writing computer programs. You could probably apply quantitative techniques for analyzing the latter to the former.

Human language, even more formal ones like legal writing, has enough ambiguity in concepts (justice, love, fairness, etc.) and wording (theology, philosophy, or legal texts) that it cannot be compared to programming language whose semantics are constrained to running on a well-specified machine.

Legal writing is different, though, and is more amenable to this.

Much like ethereum smart contracts, but for legal treatises.

Many (most?) bilateral treaties are based on a model. Counsel will keep the model up to date so instances of the treaty require less work. This task is often handed to interns (of which I was one 25 years ago).

True, most non-lawyers are surprised that stuff is copy and pasted, but as clockwork_prior says downstream, there's even an organization (ALEC) that encourages copy and paste with state laws, and the UCC is also a good source for copy-and-paste for laws. It's standard practice. Actually most commercial contracts including leases are copy and paste. The trick during negotiations is to slip a customized paragraph into the boilerplate and try and pull a 'fast one' on your opposition. For consumer contracts they sometimes insist on different font (bold) and font size for controversial passages like waiving rights.

Well, ALEC certainly supports that model - 'Missouri Governor Jay Nixon has vetoed a bill that included a drafting error copied-and-pasted from American Legislative Exchange Council (ALEC) model legislation, and criticized ALEC members for having "simply parroted ... the ALEC model act without alteration."

"While some may believe that such an error is 'close enough' for a model act, it cannot be allowed to become the law of this State," Nixon, a Democrat, wrote in his veto message. "Particularly in an area of the law that is the subject of ongoing litigation, a glaring defect such as this cannot be ignored."

Missouri Senate Bill 508, sponsored by ALEC member Sen. Mike Parsons, was modeled after the ALEC Navigator Background Check Act, and aimed to create new hurdles for health care “navigators” who enroll Missourians in healthcare plans under the Affordable Care Act. The bill was voted on by ALEC's corporate and politician members at ALEC's December 2013 meeting, and became official ALEC policy in January 2014.

The original ALEC model legislation would have required a navigator to submit fingerprints to the state highway patrol for a background check, and referenced the federal Public Law 92-554 -- but the correct reference is to Public Law 92-544, which deals with federal criminal records. The error in the ALEC model bill was copied-and-pasted verbatim into the Missouri legislation.'

'we reveal that most PTAs copy a sizable majority of their content word for word from an earlier agreement'

I believe the EU term of art for this is 'harmonization.' Surprising that it takes the excellent Kevin Lewis to bring up an interesting perspective of Brexit, since the British need to remove so much more or less 'copy and pasted EU law before actually leaving the EU.

"tomato imports will pay 5% tariff".

"tomato imports will pay 55% tariff"

Both sentences are very similar, the copy and paste is evident. However, the consequences are drastically different. Words are tools to achieve a goal, not the goal itself.

Your example is trivially absurd. The important parts of those contractual terms are the meanings of "tomato imports," "will pay," "tariff," and % of what number. The legal consistency and mutual understanding and agreement of those terms is what matters. The 5 or 55 is a customized detail. Clearly typos matter, but that isnt the issue here.

A better example may be:

"imports of tomatos that are red will pay 5% tariff"
"imports of tomatos which are red will pay 5% tariff"

I agree: typo-catching is not the real trick here.

Good example. But while the meaning of these sentences is clear from the rules of English grammar, the law tends to give effect to every part of a contract. To construe "which are red" as merely descriptive of most tomatoes would give it no effect. I'm not terribly conversant with contract law, but I suspect that if the the intention was to pay only for red tomatoes, the "which" clause would still have the effect of a requirement.

a. An interpretation that gives a reasonable, lawful, and effective meaning to all the terms is preferred to an interpretation that leaves a part of the contract unreasonable, unlawful, or of no effect. (Rest. 2d § 203(a).)

The test of admissibility of extrinsic evidence to explain the meaning of a written instrument is not whether the instrument appears to be plain and unambiguous on its face, but whether the offered evidence is relevant to prove a meaning to which the language of the instrument is reasonably susceptible (see Pacific Gas & Electric Co. v. G. W. Thomas Drayage & Rigging Co.).

The party that seeks to have a contract term interpreted in a narrower sense that is more favorable to him bears a substantial burden of proof. (See Frigaliment Importing Co. v. B.N.S.Int’I Sales Corp.)

Where the parties have attached different meanings to an agreement, it is interpreted in accordance with the meaning attached by one of them if at the time the agreement was made that party did not know or have reason to know of any different meaning attached by the other, and the other knew or had reason to know the meaning attached by the first party.

"Such copying and pasting is most prevalent among low-capacity governments that lean heavily on existing templates, and powerful states that desire to spread their preferred rules globally."

Capacity, in these contexts, seems more important relative to state economy size.

The US, for ex, has more absolute capacity to Switzerland or Iceland or Singapore. But at the same time, the complexity of designing a treaty that is to the advantage of all US interests is much higher. Increased capacity by education, productivity etc seems more clearly advantageous than increasing capacity via state size.

Copying should also be more prevalent in deals states which are learning and copying to develop economies, rather than frontier economies (If X treaty provision worked for Y, why not Z? So copy).

Copying of rules by large powerful states also tends to coincide with slowness by large states to conduct trade deals - the EU is generally much slower at conducting external deals than proximate non-EU states. The perils of trying to create a large standard issuer in opposition to allowing standards to "emerge" through bilateral deals between small states and from consultation with nongovernment aligned international bodies.

Most of the law that governs commercial relations in America is determined by the states. That initially resulted in very different laws in the different states subject to very different interpretations, which discouraged commercial relations across state lines. In the later 19th and early 20th centuries, the idea developed to create model acts that could be adopted by the states. States could adopt changes but the overall framework would be the same. Today, almost all of the states have adopted these model actss, from the model business corporations act to the uniform commercial code, which have greatly facilitated commercial relations among the states.

Today, something similar is happening with state laws that govern various occupations, such as the practice of medicine. Expansion of telemedicine has been slow in part because of 50 different state laws that govern the practice of medicine: a physician cannot "practice medicine" in a state without getting a license in the state. Several years ago a model telemedicine act was developed and presented to the states. The model act has a reciprocity provision: if a state allows physicians located in another state to practice telemedicine in the state (a physician need not be physically present in the state to be considered as engaging in the "practice of medicine" in the state), then the other state will do the same. States have been slow to adopt the model telemedicine act in part because many patients are located in one state (Florida) while many physicians are located in another (Massachusetts), and the physicians in Florida aren't keen on letting physicians in Massachusetts "practice medicine" in Florida.

Trade agreements are meant to facilitate trade between countries, and the same considerations apply: model treaties reduce the differences that can produce different interpretations or ambiguity and thereby discourage trade.

Open Skies treaties are mostly based on a US template that doesn't vary much across instances. Tax treaties are mostly based on a model but with much greater variance.

The double taxation treaties are usually based either on an old US template or on the OECD's model tax convention, with the OECD template becoming more common. With thousands of country dyads the OECD template makes sense to the negotiators most of the time. But the recent version of it includes BEPS, which seems to be a concept resisted by US exceptionalism.

Judicial decisons are also usually a product of copy and paste.

True, but usually that comes from citing precedent. Treaties have a lesser need to do so. However, a treaty is in effect a contract, and prior contracts with successfully clear wording are a good model for future contracts.

This is why the US Constitution was modelled after colonial constitutions, and why foreign constitutions were based on the US constitution.

Almost everything in law is copied and pasted. Judges often copy facts and arguments from the prevailing party verbatim. This isnt just for ease but to demonstrate they relied on the pleadings.

Union labor contracts as well.

In software:

a software design pattern is a general, reusable solution to a commonly occurring problem within a given context in software design. It is not a finished design that can be transformed directly into source or machine code. It is a description or template for how to solve a problem that can be used in many different situations. Design patterns are formalized best practices that the programmer can use to solve common problems when designing an application or system.

It would not surprise me at all that treaty negotiators have patterns of their own.


This is a trivial finding. International agreements are contracts. Parties tend to use contract terms that are mutually understandable, agreeable, monitorable, and enforceable.

Diplomats and their staffs are also notoriously lazy, risk averse, and uninventive. They hew close to well established norms.

This is not surprising.

Parts of the Bible were cut and pasted as well.

+1 hilarious , on a Mac or PC

Thank you, Father, for I have sinned....

The more interesting question is whether anyone involved actually knows and applies all the copy and pasted content.

In Labor contracts, a means of applying pressure on the employer short of a strike is to 'work to rule'. In other words follow the labor agreement to the letter, which makes things grind to a halt.

If, along with signing the agreement, there are not funds and arrangements for enforcement and monitoring adequate to the trade flows the agreement is useless and meant to be ignored.

This makes a lot of sense. After all, if Nolo Press and Legalzoom do this for just about every type of legal procedure, why can't treaties take basically the same approach?

Of course. There are ‘genealogies’ of FTAs/PTAs etc depending from early US models and European models. You can trace them in the WTO’s RTAIS database. Then, more complex & comprehensive regional agreements recapitulate a lot of GATT/WTO, too. No one needs to re-invent the technology from the ground-up.

Australia’s first FTA (other than the sun-generis mid-90s agreement with NZ) was the AU-USA FTA (2004) that closely followed the model of the USA-SGP agreement negotiated immediately before. AU’s subsequent agreements with SGP, THA etc also worked from that model.

Still, the claimed “95%” similarity, seems high. Also “copy-paste” of whole chapters, in my experience, happens only in the case of procedural stuff (accession, entry into force) and standard exceptions and exclusions of coverage (e.g. non-commercial services; preferences applying to indigenous populations etc) in my experience. And even then, not always.

One practical reason for this parsimony (other than the abbreviation of post-signature legal checking) is to minimise the complexity for “stakeholders”: a sort of anti-spaghetti-bowl safeguard that allows businesses and bureaucrats to more easily navigate the confusing network of agreements, many of which overlap. Then, the schedules of concessions attached to these agreements — the specified preferences in the customs tariffs and laws affecting access to services markets — are also very similar or identical in each country’s agreements. Once we have made extensive “concessions” to the USA and to CHN why wouldn’t we do the same for our TPP partners and the EU?

Comments for this post are closed