The Power of Attorney: How Can One Person Legally Act for Another?

Somewhere in nearly every family's filing cabinet, or increasingly in a cloud folder, sits a document that quietly does something remarkable: it lets one human being act, with legal effect, as though they were someone else. Sign their name. Sell their house. Withdraw from their bank account. Make decisions about their medical care when they cannot speak for themselves. The power of attorney looks unassuming — often just a page or two of boilerplate — but it rests on one of the oldest and most carefully bounded doctrines in Anglo-American law: agency.

The document raises an obvious question the moment you look closely at it. If legal identity and legal responsibility are supposed to track the individual person, how can that person hand their own legal voice to someone else, and under what conditions does the law still consider the resulting act genuinely theirs?

An Old Answer to a Practical Problem

Agency law did not begin with powers of attorney; it began with merchants, stewards, and factors who needed someone to act on their behalf while they were elsewhere. English common law developed the law of principal and agent to answer a basic commercial need: a merchant trading overseas could not personally negotiate every transaction, so the law had to decide when an agent's handshake bound the principal who was never in the room. The maxim that emerged, qui facit per alium facit per se — he who acts through another acts himself — became the doctrinal backbone of delegated authority across commercial and personal law alike.

The power of attorney itself is simply the written instrument that evidences this agency relationship for third parties who were not present when the principal granted it. A bank, a title company, or a hospital cannot read minds; they need a document establishing that the person standing in front of them really does have authority to act for someone else. The written instrument, often further authenticated by a notary, performs that evidentiary function, much as the broader inquiry into seals, stamps, and symbols of authority explores for other instruments that must prove themselves to strangers.

The Instrument's Structure

Modern power of attorney law in the United States is primarily a matter of state statute, and most states have adopted some version of the Uniform Power of Attorney Act, drafted by the Uniform Law Commission to standardize the form, durability, and acceptance requirements of these instruments. A power of attorney can be general, granting broad authority over the principal's financial affairs, or limited to a specific transaction — closing a single real estate sale, for instance, or representing someone at a single meeting.

A distinct category, the durable power of attorney, is drafted to survive the principal's subsequent incapacity, a departure from the traditional common-law rule that an agent's authority automatically terminated the moment the principal became incompetent. Nearly every state now allows for durability, generally by including specific statutory language in the document itself, precisely because families needed a mechanism to manage a relative's affairs through illness or cognitive decline without resorting to a court-supervised guardianship or conservatorship proceeding.

A related but legally distinct instrument, the healthcare power of attorney or healthcare proxy, authorizes someone to make medical decisions rather than financial ones, and is typically governed by separate state healthcare-decision-making statutes rather than the general power of attorney act. The two instruments are often prepared together but serve different gatekeepers — financial institutions on one hand, hospitals and physicians on the other — each with their own acceptance practices.

What the Agent Actually Owes

The authority granted by a power of attorney is not a blank check. The agent, often called an "attorney-in-fact" in this context (a term with no necessary connection to a licensed attorney at law), becomes a fiduciary the moment they accept the appointment. Fiduciary duties under agency law generally include a duty of loyalty, a duty to act within the scope of the authority actually granted, a duty to keep the principal's property separate from the agent's own, and a duty to act in the principal's best interest rather than the agent's. An agent who exceeds the document's actual scope, or who uses the position for self-dealing, is acting ultra vires — beyond the granted power — and the principal, or after death the principal's estate, generally retains remedies against that agent even though the document itself looked facially valid.

This is the point where the instrument's apparent simplicity becomes deceptive. A power of attorney authorizes the agent to act for the principal within defined limits; it does not dissolve the principal's own underlying rights and obligations, nor does it make the agent personally the party to every contract signed. When an agent signs correctly — as agent, on behalf of the named principal, within the granted scope — the resulting contract binds the principal, not the agent personally. When those conditions are not met, the clean legal fiction can unravel in ways that matter enormously to whoever is relying on the signature.

Case Files: Where Courts Draw the Line

Courts have repeatedly had to decide what happens when third parties relied on a power of attorney that, for one reason or another, was no longer valid. The traditional common-law rule was that an agent's authority terminates automatically on the principal's death, incapacity (absent durability provisions), revocation, or the expiration of the stated term — and that termination can occur before a third party has any way of knowing about it. This created obvious risk for banks, buyers, and other parties who relied in good faith on a document that looked current.

Statutory reforms, including provisions found in the Uniform Power of Attorney Act, have addressed this by protecting good-faith third parties who accept an acknowledged power of attorney without actual knowledge that it had been revoked or had otherwise terminated, and by allowing agents or principals to execute affidavits certifying the instrument's continued validity. These protections do not, however, validate a power of attorney that was invalid from the outset — for example, one executed after the principal had already lost the capacity to grant it, or one procured through fraud or undue influence. Capacity at the moment of execution remains a threshold requirement that acceptance statutes cannot retroactively cure.

The Limits Nobody Can Waive Away

A persistent source of confusion is the belief that a sufficiently broad or creatively worded power of attorney can authorize an agent to do anything, including acts the principal themselves could not lawfully perform, or acts requiring an independent license or judicial order. It cannot. An agent cannot use a power of attorney to make a will for the principal in most states, cannot generally change beneficiary designations absent express authorization, and cannot exercise authority the document itself does not actually grant, regardless of how expansively the agent or a third party chooses to interpret vague language. Courts construe powers of attorney according to the actual text, and ambiguous grants of authority are frequently read narrowly rather than broadly, particularly where self-dealing is alleged.

Nor does a power of attorney substitute for the kind of delegated institutional authority discussed in the inquiry into quo warranto and an official's authority to act. A private agency relationship between two individuals operates under contract and fiduciary law; it does not confer public office, governmental power, or any authority beyond what the principal themselves actually possessed and chose to delegate. The document cannot grant more authority than its signer had to give.

Springing Powers and the Timing Problem

Some principals prefer a "springing" power of attorney, one that only takes effect upon a triggering event such as a physician's certification of incapacity, rather than immediately upon signing. The arrangement sounds protective, but it introduces its own friction: a bank or title company must now evaluate whether the triggering condition has actually occurred, often based on a doctor's letter it has no independent way to verify. The Uniform Power of Attorney Act and many state statutes address this by allowing the document to specify exactly how the triggering event is proven, but the underlying tension between protecting the principal from premature use and ensuring the agent can act when genuinely needed has no perfect solution, only a negotiated one spelled out in the drafting.

What Survives

What endures from centuries of agency doctrine is a workable, carefully fenced answer to an old problem: one person can act for another, with real legal consequences, because the law has built a structure of accountability around that delegation rather than simply trusting it on faith. The power of attorney is powerful precisely because it is limited — bounded by scope, by fiduciary duty, by the principal's own underlying capacity, and by the third party's reasonable reliance. Strip away any one of those limits, and what remains is not expanded authority but an instrument with no legal foundation at all.

Sources & Further Inquiry

• Cornell LII Wex: Power of Attorney — overview of agency concepts underlying the instrument.
• Uniform Law Commission: Power of Attorney Act — the uniform statutory framework adopted by most states.
• Cornell LII Wex: Agency — fiduciary duties and scope-of-authority doctrine.
• Cornell LII Wex: Fiduciary Duty — loyalty and care obligations owed by agents.
• Internet Archive — digitized state probate and agency-law treatises for historical comparison.

Continue the Inquiry

The document may be only the beginning. Behind every instrument lies a question of authority, obligation, and proof. Continue the investigation through Mil.L.

Explore Mil.L. →