Rhode Island to Improve Interpreter Services After ACLU Complaint
Rhode Island laid off interpreters to save money and spent the next several years dealing with the consequences. The state agreed to improve its language services after a complaint from the state chapter of the American Civil Liberties Union, filed in 2007 when the Department of Human Services cut all of its Southeast Asian language interpreters and one of its two Portuguese interpreters.
What the cuts actually removed
The interpreters who went were not a convenience. They were the mechanism by which people who spoke little or no English applied for food assistance, medical coverage and disability support. Remove them and the application process becomes a wall. The benefit still exists on paper. The person cannot reach it.
Rhode Island's demographics made the cut particularly consequential. The state has substantial Portuguese-speaking and Southeast Asian communities, many of them elderly, many of them exactly the population that human services exists to serve. Cutting the interpreters who served them was, in effect, a benefit reduction targeted by language.
The legal basis for the complaint
The ACLU's argument did not rest on fairness in the abstract. Federal law obliges agencies that receive federal funding to provide meaningful access to their programmes for people with limited english proficiency, and the federal guidance collected at lep.gov sets out what that requires. An agency that dispenses federally funded assistance and cannot communicate with a portion of its applicants is not delivering the programme it is funded to deliver.
The resolution came as an agreement between the state's Department of Human Services and the federal government to improve assistance for people who speak little or no English. Agreements of that kind typically require a written language access plan, staff training, and provision of interpretation at every point where a decision is made.
The false economy
The reason budget cuts land on interpreters so often is that the saving is easy to calculate and the cost is not. A salaried interpreter is a visible line item. The consequences of removing them are diffuse: applications abandoned, appointments repeated, errors made and later corrected, appeals filed, and eventually a legal complaint with attached remediation costs.
Health systems that have studied this arrive at the same conclusion. Language services are cheap relative to what happens without them. The pattern recurs in courts, in schools and in social services, and it recurs because the savings are booked in one department and the costs land in another.
What good language access looks like
Agencies that do this competently share a common structure:
- A written language access plan naming the languages served and the trigger for providing an interpreter
- Trained staff who know how to request one, rather than staff who ask the client to bring a relative
- Telephone and video interpreter services for languages too rare to justify a salaried post
- Vital documents translated in advance rather than sight-translated at the counter
- Data collection on which languages are actually being requested, so provision follows demand
The reliance on family members deserves a separate warning. Using a client's child to interpret a benefits interview or a medical consultation is common, free and indefensible. The child lacks the vocabulary, the adult loses their privacy, and the agency has no way to verify that anything was conveyed accurately.
Ad hoc provision is the real risk
The failure mode in most public agencies is not an outright refusal to provide interpretation. It is provision that depends on who is on shift. A bilingual caseworker becomes the department's unofficial interpreter, doing a second job for no extra pay and no training, until they leave. Then the capacity leaves with them.
Professional interpreters have documented this for years, and it comes up constantly in practitioner communities such as r/interpreters. Bilingualism is not interpreting. Interpreting is a trained skill with ethics, protocols and a duty of accuracy that a helpful colleague has never been taught.
How agencies get the language list wrong
A recurring technical failure sits underneath these disputes: agencies guess at which languages they need to serve instead of measuring it. The census gives a rough picture of a state, but it does not tell a specific benefits office which languages walk through its door on a Tuesday morning, and the two can differ sharply in a city with a concentrated refugee population.
The fix is unglamorous. Record the language of every request for assistance, including the ones that were turned away or handled badly, and review the data quarterly. Agencies that do this discover two things almost immediately. The first is that a handful of languages account for most of the demand, which makes provision cheaper than they feared. The second is that a long tail of rare languages appears a few times a year, and those are the cases where a telephone service is the only sane answer, because no one is going to employ a salaried Karen or Kinyarwanda interpreter for eleven appointments.
What happens when the plan is written and then ignored
A language access plan that lives in a policy binder changes nothing. The agencies where provision actually works are the ones that made the interpreter request part of the workflow, so that a caseworker opening a file for a client flagged as needing Portuguese sees the interpreter booking step before they can proceed. Compliance that depends on a busy person remembering an obligation is compliance that fails on the first busy day.
Why the case still matters
Rhode Island's agreement is a small item in the history of civil rights enforcement, and a useful one, because it establishes the ordinary version of the principle. Language access is not a courtesy extended to newcomers. It is the condition under which a public service is actually available to the public that funds it.
Every few years another state discovers this by cutting the interpreters first and reading the statute afterwards.