The Agent at Your Kitchen Table

Doris ran a school district for eleven years. Before that she was a principal, before that a classroom teacher, before that a graduate student who wrote her master’s thesis on organizational systems. She is 73, lives alone in the house she and her husband bought in 1987, and is not confused about anything. She is one of the most administratively capable people I have invented for this series, which is why I am using her to make this point.

On her kitchen table this morning: a Medicare Summary Notice she has read three times and cannot parse, a property tax bill she suspects has an exemption attached to it but cannot confirm, a prescription that cost $340 last month for a drug she has been told may have a manufacturer discount program she has never located, and a letter from her pension administrator requiring a response by a date that passed eleven days ago. She will spend four hours on hold this week. She will resolve two of these four things. The third will slip. The fourth will become a problem in March.

Doris is not failing. The task is impossible. This is an important distinction, and the rest of this piece depends on holding onto it.

The complexity that nobody named

Blue Gray Matters has spent considerable space on the cost architecture of aging in America: what things cost, why they cost that much, who can afford them, who cannot. This installment is about something different. Not cost architecture. Complexity architecture.

The average older adult managing their own affairs interacts with somewhere between fifty and one hundred distinct administrative systems in a given year. Medicare and a supplemental plan. Part D. Social Security. A pension administrator. A bank. An investment account. A property tax authority. A utility provider. A homeowners insurance carrier. A life insurance carrier. A doctor’s billing office, usually three or four of them. A pharmacy. An internet provider. A local senior services organization they contacted once and now receive quarterly newsletters from. Each system sends communications. Each assumes periodic engagement. Each has its own portal, its own password requirement, its own procedures for disputing a charge or updating an address or appealing a decision. Each was designed by people who thought about their system, not about all the other systems the same person is managing at the same time.

The class dimension of this is not incidental. Wealthy people have accountants, financial advisors, estate attorneys, and personal assistants who manage these systems on their behalf. Not all of them, but enough. The burden shifts to professionals who are paid to absorb it. Middle-class people manage it themselves, with shrinking margins as the volume increases and as age, health, or grief occasionally interrupts their capacity to stay on top of it. Poor people face the most paperwork of all, because every benefit program is a separate bureaucracy with its own eligibility determination, its own renewal cycle, its own documentation requirements. The administrative load of poverty in America is staggering and largely invisible in policy discussions that focus on dollar amounts rather than hours of human life consumed.

Pamela Herd and Donald Moynihan documented this pattern in their 2019 book on administrative burden: that burden functions as a tool of exclusion, whether by design or by indifference, and that its costs fall most heavily on the people with the least capacity to absorb them. This is not a technology problem. But technology is where the most plausible near-term relief lives, and that is what this piece is about.

Why another app is not the answer

The first response to complexity is usually to add a layer that manages the complexity. And in the technology space, that response has produced dozens of apps designed to help seniors track benefits, compare insurance plans, find local services, and organize their documents. They are not useless. They are also not sufficient, and understanding why clarifies what actually would be.

Every app is another system. Another login. Another password. Another interface to learn, another notification to manage, another entry point that assumes the user has the bandwidth to engage with it deliberately. Apps are reactive tools: they answer when asked, they retrieve information when prompted, they send reminders to people who have already decided to pay attention. What they require, in every case, is that the person experiencing the problem knows they have the problem, knows to ask, and has the cognitive and logistical capacity to use the tool at the moment they need it.

This is exactly the capacity that the administrative cascade erodes. The person who has just received a confusing Medicare notice is not in the optimal condition to learn a new app that might help them decode it. The person managing a medication change, a housing decision, and a pension letter in the same week is not the person who benefits most from a tool that requires deliberate setup and engagement. Reactive tools work for people with surplus capacity. The people this series is writing about often have none.

What would actually help is something categorically different: a system that perceives gaps without being asked. That notices the Medicare Summary Notice arrived and flags the appeal window before it closes. That checks the property tax record against the exemptions available in that jurisdiction. That identifies the manufacturer assistance program for that specific drug and generates the application. That does not wait to be prompted because the whole point is that the person managing seven things at once may not know to prompt it.

This is not reactive assistance. It is proactive representation. The distinction is fundamental.

The asymmetry that is already here

The reason proactive representation matters is that the other side of the table already has it.

When you file a claim with your insurance company, that claim is reviewed by a system optimized to minimize the company’s payout. Not by a malicious human. By an algorithm trained on the company’s interests, which are not your interests. Pharmaceutical benefit managers, the intermediaries who negotiate between drug manufacturers and insurance plans, deploy their own algorithms to determine which medications get favorable placement in your plan’s formulary and which require prior authorization. Your bank’s fee structure, your credit card’s default settings, your utility company’s billing practices: these are all the products of systems designed with institutional interests at their center. This is legal, common, and not especially controversial in business terms.

On your side of the table: a phone, a hold time, and whatever knowledge you have accumulated about a system designed by someone who assumes you will not challenge it.

This asymmetry has always existed. What is changing is the scale and the speed. Every significant industry touching older adults is deploying AI agents to manage their side of the equation more efficiently. Insurance claim reviews that once took human adjusters now happen in seconds. Prior authorization systems that once required a phone call now operate through automated portals. The institutional AI is already deployed. The individual equivalent is not.

A buying agent, a benefits navigator, an administrative advocate that monitors your situation and acts on your behalf: this is not a luxury technology for the affluent. It is a corrective to a power imbalance that is becoming structural. The reason it does not yet exist at scale for people who need it most is that the people who need it most cannot pay for it, and the business model for building it has not been solved. The most honest framing of that problem: employers, health systems, and government agencies should pay for this because a person who loses their housing because they missed a Medicaid renewal form, or who goes without medication because they didn’t know about the assistance program, or who pays a penalty for a late pension response, costs those systems far more than the intervention would have.

That argument has not yet won in the rooms where it needs to win.

What you can do this week

Three phone numbers. Not apps. Not portals. Phone numbers staffed by people whose job is to help you with exactly the things on Doris’s kitchen table.

The first is your local Area Agency on Aging. There are 622 of them across the country, established under the Older Americans Act, and most people have never heard of them. They provide benefits counseling, caregiver support, transportation assistance, meal programs, and connections to local services. You can find yours by calling the Eldercare Locator at 1-800-677-1116 or by visiting eldercare.acl.gov.

The second is your State Health Insurance Assistance Program, known as SHIP. Every state has one. SHIP counselors provide free, unbiased help with Medicare questions, including plan comparisons, billing disputes, appeals, and the supplemental insurance options that most people are never adequately counseled about when they first enroll. A SHIP counselor is not selling anything. They are paid by the state to help you. In an average counseling session, SHIP counselors identify savings or benefits of several hundred to several thousand dollars per person. The number for your state’s SHIP is available at shiphelp.org or through the same Eldercare Locator.

The third is 211, the national social services helpline available in most parts of the country. If you don’t know what assistance programs exist in your area for utilities, food, transportation, or housing, a 211 operator can navigate that landscape more quickly than you can from a search engine.

These are not technology solutions. They are human solutions that already exist, are already funded, and are systematically underused because the people who need them don’t know to call.

Doris made two calls on Thursday afternoon. SHIP identified a Medigap plan change that will save her approximately $1,800 this year. The property tax office confirmed she qualifies for an exemption worth $600 annually; the previous owner had it, and it lapsed when ownership transferred. The pharmacist, once she asked specifically, knew the manufacturer’s assistance program and helped her apply. The prescription will cost $35 next month.

Three calls. About ninety minutes of her time. Roughly $2,400 in annual savings she would not have found without knowing to ask.

She shouldn’t have needed to know to ask. Someday, something will monitor her situation continuously and make those calls on her behalf, and the result will be the same, except the ninety minutes will be spent on something else. That day is not fully here. The three phone numbers are. So is the problem on her kitchen table.

Related reading: BGM-1A (The Price Tag No One Shows You), BGM-1B (The Fine Print They Hand You at 65), BGM-1C (The Pharmacy Trap), BGM-7B (Social Security: The Timing Game), BGM-7D (The Long-Term Care Conversation), BGM-11A (Two Hip Fractures)

Blue Gray Matters is an independent publication. We have no financial relationship with any product, device, or service mentioned here.