Judicial review asks us to accept an arrangement that, described plainly, looks like a democratic paradox: a small number of unelected judges, appointed rather than voted into office, may strike down a law passed by the elected representatives of the people who will live under it. Stated this starkly, the practice seems to invert the basic logic of self-government, substituting the judgment of a handful of lawyers in robes for the judgment of everyone else. And yet the practice has proven durable across a wide range of democracies, adopted and retained by legislatures that could, in principle, simply abolish it. That durability suggests the paradox, however real it looks on the surface, is not quite the whole story.

The strongest defence of judicial review begins by noticing that a legislature represents only the current majority, elected under today's conditions and answerable to today's passions, while a constitution is typically written to bind not only today's majority but every future one as well. When a legislature, under considerable public pressure during a moment of crisis, passes a law curtailing the speech or assembly rights of an unpopular minority, it is acting entirely within its ordinary democratic mandate -- it was elected, after all, and it is doing what a considerable share of the electorate currently wants. Judicial review exists precisely to check this kind of majoritarian excess, treating certain rights as promises the political community has made to itself in a calmer moment, promises that a later, more agitated majority should not be able to break simply because it currently holds the votes to do so.

This raises an obvious rejoinder: judges are unelected too, and no less fallible than legislators, so why should their judgment prevail over a democratically elected body's own? The honest answer is that, in most systems, it does not prevail absolutely. A legislature that disagrees strongly enough with a court's reading of a constitution can, through the deliberately harder process of constitutional amendment, overrule the court entirely; judges themselves are typically appointed through processes -- nomination by an elected executive, confirmation by an elected legislature -- that connect the judiciary to democratic accountability, only at one remove and on a considerably slower timescale than an ordinary election. Judicial review, properly understood, is not a permanent transfer of final authority away from the people; it is a deliberately slower, harder channel through which the people may still act, precisely so that changing a foundational promise takes more than the ordinary majority that can pass an ordinary law.

The practical effect of this arrangement extends well beyond the relatively rare cases a court actually decides. Legislators drafting a bill they expect to face judicial scrutiny tend to draft it with that scrutiny already in mind, narrowing an overbroad provision or adding a procedural safeguard not because any court has yet objected, but because the credible possibility of a future objection shapes the bill before it is ever introduced. This shadow effect means judicial review need not invalidate very many laws at all to considerably influence how laws are written in the first place, since anticipating a court's scrutiny becomes, for a competent legislative drafter, simply part of writing the bill well.

Judicial review, then, resolves its own apparent paradox not by claiming judges possess superior wisdom, but by accepting a deliberate trade-off: certain rights are made harder, though never impossible, for an ordinary majority to erode on an ordinary timescale, in exchange for a slower and more considered process before any such erosion can occur. This is neither pure majoritarianism, in which whatever the current electorate wants automatically becomes law, nor pure judicial supremacy, in which judges hold final and unreviewable authority; it is engineered friction, placed deliberately between a passing majority and a constitutional promise, so that only a majority considerable and persistent enough to survive the friction can actually change what the promise protects.