A Day With No Spare Place

A free hour on a calendar can be a very crowded place.

It may hold the trip you postponed, a call that must happen during office hours, the meal you have not prepared and the chance to sit quietly. The fact that none of these is written down does not make the hour unclaimed. If someone asks whether you are available, the answer depends on which claim you are willing to disappoint.

Care adds another complication: some work must happen at a certain time, and some must happen whenever the need arises. A ride to an appointment can be scheduled. Waiting for a service to confirm its arrival may leave a much larger part of the day unsettled.

To understand the strain, we need more than a count of minutes. But minutes still matter. They put a limit on what a well-meaning speech can ask a body to do.

Let us look closely at one good source of them.

What a diary can tell us

The U.S. Bureau of Labor Statistics asks people to describe the previous day in the American Time Use Survey, or ATUS. The diary follows activities across a full day. Additional questions identify unpaid eldercare and activities during which it happened. For the published eldercare estimates, a provider has at least one recipient aged 65 or older who needs assistance because of an aging-related condition.¹

This is a narrower population than the broad caregiving survey in Chapter 1. ATUS does not stand for every form of disability support or care for adults of every age. Its published population also begins at age 15. Different definitions produce different totals, so placing those totals next to each other does not create a trend.

BLS estimated 38.2 million unpaid eldercare providers in 2023–2024. On a given day, about 28 percent provided care. Those who did spent an average of 3.9 hours doing so.¹

Read the last two sentences together. The 3.9 hours apply to days when care was provided. They do not mean every provider did nearly four hours every day. A person may count as a provider because they gave help during recent months, yet have no care activity on the particular diary day.

Now reverse the mistake. If we average across all provider-days, including days with no recorded care, the number becomes much smaller. That does not mean the people actively giving care worked only the smaller number of hours.

Both averages are legitimate. They answer different questions.

Imagine a tiny example with ten equally weighted diary days. Care occurs on two, for three hours each. There are six care hours in total. The average across all ten days is 0.6 hours. The average across the two care days is three hours. No contradiction exists. The difference comes from the people and days included below the division line.

That line is called the denominator. In a life crowded with responsibilities, it can decide whether a statistic explains the situation or makes it disappear.

An original look at the public data

For this book, we analyzed the public ATUS files for 2023 and 2024. First we checked that the code reproduced the published eldercare benchmarks using BLS’s original age range. It did at the published rounding, including the estimate of 7.616 million providers with their own children under 18 living at home and the finding that 86 percent of those parents were employed.

We then limited our comparison to adults. After excluding six providers whose diary-day care answer was missing, the analysis included 2,528 people. We used the survey’s weights so that the sample represented the population it was designed to describe, and its replicate weights to estimate sampling uncertainty. The technical appendix records the definitions, checks and calculations.²

The first plan was to compare employment, resident children and care frequency together. The sample could not support all those subdivisions well. Some groups had too few care days to justify a confident-looking estimate. We narrowed the question instead of decorating thin results with extra decimal places.

The figure shows three broad groups. “No own resident child under 18” does not mean a person has no children or no other dependents. “Not employed” includes people in different circumstances, including retirement. These are descriptions at the time of the survey, not explanations for why a person’s day took its shape.

Read the full figure description, including values, uncertainty intervals and sample sizes.

Among employed providers with their own resident child under 18, eldercare averaged about 0.37 hours across all provider-days and 2.44 hours on care days. That is roughly 22 minutes across all days, or two hours and 26 minutes on care days. The corresponding estimates for employed providers without such a child were 0.71 and 2.94 hours. Among providers who were not employed, they were 1.90 and 4.83 hours.

Those are estimates, not exact population facts. The care-day estimate for employed resident parents is based on 53 diaries and is especially uncertain: its 95 percent interval runs from about 1.40 to 3.48 hours. The data do not justify a confident ranking of care-day duration between the two employed groups.

The broader lesson is already visible. An all-days average can make an intermittent role look small. The care-day average reveals a substantial block of time when that work occurs. Neither tells us how easy that block is to fit around the rest of life.

A lower care count can sit inside a crowded day

The employed resident-parent group recorded less eldercare time on average than some other groups. It would be a mistake to conclude that these caregivers had the easiest lives.

They were also parents, and paid work occupied part of their days. In the same analysis, their recorded leisure and sports, including related travel, averaged about 3.49 hours a day, compared with 4.36 among employed providers without their own resident child under 18. These are averages across all days of the week. They are not the length of an uninterrupted evening break, and they do not establish that caregiving caused the difference.²

Age, work schedules, health, the needs of the person receiving care and other differences can shape the comparison. People may also change employment because of care. A cross-section of diary days cannot untangle all those directions.

Even so, the result corrects a bad habit of reasoning: treating the amount of one duty as a complete measure of how much a person carries. Someone with fewer recorded eldercare hours can have less room to absorb an extra appointment. Someone who is not employed can provide many more hours without having an easy or freely chosen schedule.

There is no prize for winning a contest of exhaustion. The point of comparison is to understand which resources a person needs, not to rank who deserves sympathy.

This matters inside families. A relative may say, “You only go twice a week,” while overlooking the fixed times, the travel and the other duties around them. Another may assume the retired person is always available. Employment status tells us something about a day. It does not grant the family ownership of the hours outside paid work.

Leisure does not always mean off duty

Here the data offer another useful warning. Eldercare can occur during an activity that the diary classifies as leisure, such as companionship. A person can sit watching a program with an older adult while also providing care.

If we add every eldercare hour to every leisure hour as if they were separate blocks, we can count some time twice. A colorful chart could then imply a day longer than 24 hours. The picture might look persuasive precisely because it is wrong.

There is a human error beneath the arithmetic one. Seeing someone seated can make it easy to assume they are resting. Sometimes they are. Sometimes they are present because the other person needs someone there. Sometimes the company is enjoyable and the responsibility remains real. A single label cannot settle the experience.

The analysis therefore keeps overlapping measures separate. It does not call all leisure “time off.” It also does not claim to measure every anxious thought, interrupted plan or moment of vigilance. A diary describes reported activities. It cannot fully describe what a person felt free to do instead.

Freedom to leave is a different kind of information from the activity performed while staying.

The time that cannot be moved

Picture two care tasks, each lasting an hour. One can happen at any point during the weekend. The other must happen between seven and eight on a weekday morning. The second may collide with paid work or a school journey even though the time total is identical.

Now add a third task that takes only fifteen minutes but requires the helper to remain nearby for an uncertain part of the afternoon. Counting only the active minutes will not describe the restriction. Counting the entire afternoon as active care would misdescribe it in another way.

A practical account needs both duration and timing. It should ask how much time the task uses, how fixed the time is, how much notice is given and whether another person can do it. These are questions to understand the arrangement, not a new scientific scale.

The answers can change the request. A person may not need someone to take over every Thursday. They may need transport that arrives within a reliable window. They may need appointments grouped on the same day, where the provider can reasonably arrange that. They may need an employer to allow a known shift change rather than approve emergencies one at a time.

The institution supplying the timetable has some responsibility for its effects. A service that gives a four-hour arrival window is asking a household to supply four hours of availability. Calling the actual visit brief does not remove that request.

The caregiver should not have to solve every timing problem alone. It is reasonable to ask those who control schedules to understand how their choices land in a home.

A week is more than seven copies of one day

The ATUS diary records one day per respondent. That is useful for population estimates. It is a poor basis for claiming that a particular person’s usual week has been measured.

We should especially avoid multiplying the care-day average by seven and presenting the result as a weekly care burden. A person who gives care on one or two days would be assigned hours they did not report. Even an all-days population average cannot tell us which individual experiences long stretches or sudden peaks.

Care can come in episodes. A stable month is followed by an illness, a hospital visit, a change in services. A schedule that was manageable stops being manageable. The annual average may barely hint at the week in which a job, a relationship or a person’s health came under pressure.

This is why households need a way to review change, and why services should look beyond a calm day when deciding what support is needed. Asking only what someone did yesterday can miss the coming demand. Asking only what happened in the worst week can make a temporary need appear permanent. Both accounts may be useful if their dates and purposes are clear.

Statistics help us see patterns. Planning requires us to ask what is happening here, now, and what is likely to require a new decision.

Stop looking for the unused person

When a family needs more care, it may search for the person who appears to have the most spare time. That can be a reasonable part of a conversation. It becomes unfair when appearances are treated as consent.

The person without children may have other obligations. The person working from home may have no more freedom to leave than someone at a counter. The older partner may need care themselves. A friend who helped before may be willing again, but their earlier generosity is not a standing appointment.

What matters is confirmed capacity: what someone says they can reliably do, with the resources and limits made clear. Capacity can grow when training, transport, pay or replacement care is supplied. It can shrink when health or work changes. It should be checked rather than guessed.

This brings us back to the hour on the calendar. A care plan should not be built by treating every blank as an available shift. Blanks may be where a person keeps the rest of their life alive.

The data have helped us resist a false simple story. There is no single caregiver day. The next chapter begins with one of the moments most likely to change whatever day a household had: the journey home from hospital.

Source notes