9.7 Million Direct Care Job Openings by 2034. The Government Still Has No Code for the Job.

A congressional staffer once asked me how many direct support positions were unfilled in my county. It was a fair question, asked in good faith, and I could not answer it. Neither could anyone else, and it took me a while to understand that this was not a research failure on my part.

There is no Standard Occupational Classification code for direct support professionals. They are folded into a broader aide classification, alongside people doing genuinely different work under genuinely different conditions. The Bureau of Labor Statistics therefore cannot report employment or wage data specific to the profession, because as far as the federal statistical system is concerned, the profession does not exist as a distinct thing.

I want to explain why that matters more than it sounds like it should.

What a classification code actually does

The Standard Occupational Classification system is the spine of federal labor data. It determines which jobs the Bureau of Labor Statistics counts, what wage series exist, which shortage projections get published, and what a state Medicaid agency can point to when it builds a rate.

When a job has a code, it produces a number every year without anybody having to fight for it. Employment level. Median wage. Projected growth. Regional concentration. Those numbers show up in fiscal analyses, in workforce shortage designations, in the models legislative staff use when they score a bill.

When a job has no code, none of that happens automatically. The only figures that exist are the ones somebody paid to go collect, which in this field means provider surveys with response rates and coverage gaps. Those surveys are genuinely good work. The 2024 NCI-IDD survey covers 3,936 agencies across 27 states and the District of Columbia. But a voluntary survey covering 27 states is not the same instrument as a federal data series covering all of them, and everyone in a hearing room knows the difference.

The scale of what is not being counted

PHI projects 9.7 million total direct care job openings between 2024 and 2034. That figure combines growth and replacement, which is the honest way to count openings in a field with turnover like this one.

The breakdown: home care accounts for more than 6.1 million of those openings, residential care aides about 1.1 million, and nursing assistants 664,700, even though nursing assistant employment is projected to shrink slightly. The workforce is 5.4 million people today and is projected to add over 772,000 new jobs by 2034, which PHI describes as the largest growth of any job sector in the country.

Largest growth of any sector. And a chunk of it has no occupational code.

The bill, and why I think it is worth your letter

The Recognizing the Role of Direct Support Professionals Act is S. 3211 in the Senate, sponsored by Maggie Hassan of New Hampshire and Susan Collins of Maine, with a House companion at H.R. 6137. The Senate Committee on Homeland Security and Governmental Affairs ordered it reported on August 6, 2026.

What it does is narrower than the name suggests, and I would rather you hear the limits from me than discover them later. It directs the Office of Management and Budget to consider creating a separate occupational code for direct support professionals when it next revises the classification system, and to report to Congress if it declines. OMB is scheduled to conduct that review in 2028.

Consider, not create. That is a real limitation and advocates should stop overselling it.

The Congressional Budget Office estimated the cost of the reporting requirement at under $500,000 across 2026 to 2031, and noted that OMB will likely fulfill the consideration requirement through its normal review anyway, given that more than 1,500 public submissions have already addressed direct support professionals.

The honest case against it

If you read that CBO note carefully, there is a real objection sitting in it. If OMB is going to consider this anyway during a review already scheduled for 2028, and 1,500 people have already written in, what does the statute add beyond a press release?

That is a fair question and I think it has a fair answer. The bill converts a discretionary consideration into an obligation with a reporting requirement attached. OMB can still decline. What it cannot do, if this passes, is decline quietly. Having to write down why you are not counting 5.4 million workers, and send it to a committee, is a meaningfully different thing from letting a comment docket close.

The other objection worth naming is that classification is not care. A code does not staff a shift, raise a wage, or fill a slot. Every hour spent on a data bill is an hour not spent on a rate bill. I feel that one, and I would not trade a rate increase for a code. But nobody is offering that trade, and the sequencing runs the other direction anyway. Rate arguments get scored against federal data. Build the data first and every later argument gets cheaper.

What to do this month

Check where your delegation stands. Look up S. 3211 and H.R. 6137 and see whether your senators and representative are on them. Congress.gov shows co-sponsors and updates within a day or so of any action.

Write, and ask for writing back. Name the bill by number. Note that it is bipartisan and that CBO scored it under $500,000. Ask for a written position and a date. General expressions of support cost a staffer nothing and tell you nothing.

Use the refusals as information. This is close to a free vote: bipartisan, negligible cost, no program created, no mandate on any state. If a member will not co-sponsor this, you have learned something genuinely useful about how they will treat the rate increase you plan to ask for next year. That is worth knowing before you spend a year on it.

Calendar 2028. Whatever happens to the bill, the OMB revision is coming. When that docket opens, a comment from a family describing what the job actually involves, medication administration, seizure response, behavioral de-escalation, personal care, documentation, is exactly the kind of submission that argues the job is distinct enough to need its own code.

I still cannot tell that staffer how many direct support positions are unfilled in my county. Neither can anybody else, and that is not because nobody has looked. It is because the federal statistical system has not been asked to look, and asking it to look is a thing Congress can do this year for less than the price of a single group home.