Life Sciences R&D in Puerto Rico: What Actually Qualifies and What Does Not

There is a conversation we have roughly once a month. A life sciences company with Puerto Rico operations calls, and somewhere in the first ten minutes someone says a version of this: "We do not really do R&D here, we just manufacture."

Then we walk through what their teams actually spend their time on, and by the end of the call the picture has usually changed quite a bit.

This is not a trick. It is that the word "R&D" carries a mental image, and the image is wrong. People picture discovery. Molecules on a whiteboard. Scientists in a greenfield lab inventing something that did not exist. That is research, certainly, but it is a small fraction of what regulated life sciences development actually consists of, and it is not the only thing that counts.

Start with the actual standard

Puerto Rico's current framework is Act 60-2019. The research and development activity credit lives at §3030.01. Earlier statutes, including Act 73-2008 (Legacy), still govern some grandfathered decrees on the island but are not the basis for new planning.

The model is two fixed rates. Fifty percent on direct R&D investment deployed in Puerto Rico into experimental activities. Twenty percent on indirect R&D investment deployed outside Puerto Rico or into incidental activities. That is the whole rate structure. No slider, no negotiation.

Because the rates are fixed, the entire analysis collapses into two questions. Is the activity experimental, and was it carried out in Puerto Rico. Everything else is bookkeeping.

What usually qualifies and gets missed

In our experience these are the categories life sciences companies most often leave on the table.

Process development and characterization. Determining operating ranges, understanding what happens at the edges, establishing design space, figuring out why a batch behaved differently than the model predicted. This is experimental work by definition. You are resolving technical uncertainty through systematic investigation. Companies classify it as "manufacturing support" and move on.

Analytical method development. Developing, optimizing, and validating assays. If the method did not exist in usable form and your team had to develop it, that is development. The fact that it supports a commercial product does not make it routine.

Scale up and tech transfer. Moving a process from one scale or one site to another almost never works on the first try, and the reason it does not work is technical uncertainty. The investigation required to make it work is the qualifying activity. The transfer itself is not the point. The problem solving is.

Design verification and validation for devices. Device companies are often the most surprised. Verification testing against requirements, design iteration driven by test failures, biocompatibility and sterilization development work. This is engineering under uncertainty, documented to a standard most industries would envy.

Stability and formulation work. Reformulation, excipient changes, container closure development, extractables and leachables investigation. Frequently treated as regulatory overhead. Frequently experimental.

Software supporting regulated operations. Not the license you bought. The development work your team did to build something that did not exist, where the technical approach was genuinely uncertain.

What does not qualify, and please do not claim it

This is the part that gets skipped in sales conversations, and skipping it is how companies end up with painful examinations.

Routine production is not research. Running a validated process to make product is the opposite of resolving uncertainty. Routine quality control testing against an established method is not development, even though it happens in a lab and is performed by scientists. Routine stability monitoring on an approved product is monitoring.

Regulatory submission preparation, by itself, is administrative. Market research is not technical research. Efficiency projects where the outcome was never in doubt are not experimental, no matter how much money they saved.

And the geography rule is not decorative. Work performed outside Puerto Rico does not become island research because the invoice was paid by the island entity. That is precisely what the indirect category exists to capture, at a different rate, on purpose.

The gray zone, honestly described

Most real claims live in between, and pretending otherwise is unhelpful.

A batch investigation can be routine troubleshooting or it can be genuine experimental work, depending on whether the team was following a known decision tree or actually did not know the mechanism. A validation protocol can be a checkbox exercise or it can be the documentation of a real development effort. A process improvement can be a parameter tweak or a reformulation.

The determining factor is almost always the same thing: was there real technical uncertainty at the start, and did the team resolve it through systematic investigation. If yes, you probably have a qualifying activity. If the answer was known and the work was execution, you probably do not.

Being disciplined about that line is not conservatism. It is what makes the rest of the claim credible. A claim with a clean, defensible core survives scrutiny. A claim padded with routine operations invites someone to question the whole thing, including the parts that were perfectly legitimate.

The documentation problem, which is the real problem

Here is the pattern that costs companies money.

The qualifying work happened. Scientists did real experimental investigation, solved hard problems, and generated enormous amounts of technical documentation in the process. Then at claim time, nobody can tie that work to a project definition, a cost center, and a set of hours.

The technical record exists. The financial linkage does not. And a credit claim needs both.

What good looks like is unglamorous. Projects defined at the start with a stated technical objective and the uncertainty being addressed. Time capture that distinguishes development from production support, applied by people who understand the difference. Cost allocation that ties to actual accounting systems rather than a retroactive estimate. A narrative that a technically literate reviewer who was not there can follow.

Life sciences companies have an advantage here that most industries do not. You already run document control. You already keep lab notebooks, protocols, reports, and change records to a regulated standard. The raw material for a strong claim is usually sitting in your quality system. It just was never organized for this purpose.

A practical starting point

If you want a quick read on your own situation, pull your Puerto Rico headcount and ask one question per group: does this team spend a meaningful portion of its time resolving technical questions where the answer was not known in advance.

For most regulated life sciences operations on the island, more groups say yes than management expects. Process development, analytical development, device engineering, and formulation almost always do. Validation and quality engineering sometimes do, in part. Production and routine QC usually do not, and should not be claimed.

That exercise takes a few hours and will tell you more than any estimate.

Everything above is indicative advisory context. Qualification, rates, decree terms, and administrative practice depend on your specific facts and should be confirmed before you rely on them for planning.