A huge chunk of the Supreme Court’s work lies in interpreting the statutes and regulations that govern every nook of American life. In reading statutes, the justices oscillate uneasily between two inconsistent approaches. Sometimes they distill the meaning of a disputed provision by making their best independent judgment about its structure and function. So they slap down any government officials who exceed statutory powers. Alternatively, they lament the imprecision of language, doubt their own expertise about social and political complexities, and defer to whatever reading the official gives to the statute that empowers him.

It is instructive to ask how consistently the four conservative justices — Roberts, Scalia, Thomas and Alito — and their four liberal colleagues — Stevens, Souter, Ginsburg and Breyer — apply these approaches in the battle for supremacy, with Justice Kennedy frequently the swing voter. Answer: not at all. In principle, it would nice if both sides of the ideological spectrum displayed a sound and consistent position on statutory construction. Unfortunately, each bloc is opportunistic. The litmus test for this erratic behavior boils down to a factor not found in any statute: trust.

The court’s two wings share one trait: They defer only to the government officials they trust. Otherwise, they read a statute carefully to rein in the authority of officials they don’t trust. The two factions don’t differ in their philosophy of language, or in their on-again, off-again adherence to the rule of law. Rather, the court’s liberal wing profoundly distrusts this president, but has great confidence in the domestic administrative agencies that regulate matters such as the environment. The conservative wing of the court flips over. It willingly defers to the president on national security issues while looking askance at expansionist tendencies of the administrative agencies.

Consider two key cases of this past term. Hamdan v. Rumsfeld addressed whether the Uniform Code of Military Justice authorized the president to create special military commissions to try an alleged member of al Qaeda with war crimes. These tribunals offer the accused none of the legal protections available under the UCMJ, let alone in a federal district court. The president’s power to establish these commissions unilaterally depended critically on whether he considered it “practicable” to use the standard UCMJ procedures. The liberal bloc held that a blanket presidential order for special commissions did not make any effort to show that it was impracticable to follow the UCMJ. The conservative wing wanted to defer to the president’s near-limitless discretion over foreign affairs, which courts have neither expertise nor knowledge to question.

Both wings of the court did 180-degree turns in Rapanos v. U.S. The 1972 Clean Water Act makes it unlawful to pollute “navigable waters without a permit” from the Army Corps of Engineers. “Navigable waters” are defined as “the waters of the United States, including the territorial seas.” The Corps’ original 1974 regulations stated that these waters included only those on which navigation did or easily could take place. But without explanation the Corps switched gears, so that today they embrace not only all wetlands adjacent to navigable waters — itself a stretch — but also every mudflat, slough or prairie pothole in the land. The Corps used this definition to prosecute John Rapanos, criminally and civilly, for filling in a 54-acre saturated plot of land located 11 miles from the nearest navigable river, without showing that any material from the site could reach, let alone pollute, navigable water.

The Supreme Court’s liberal bloc duly announced its deference to the Corps, while the conservatives rightly protested that such verbal alchemy could not convert land into water, nor show a family resemblance between a pothole and a territorial sea. Justice Kennedy honed in on the irrelevant by asking whether the Corps could establish some “significant nexus” between remote wetlands and navigable waters — without asking what discharges, if any, could go from one place to the other.

In surveying the landscape, note first that neither question of statutory construction is particularly difficult once the bogeyman of judicial deference is ignored. In Hamdan, Justice Stevens was right to chastise the president for flouting the Rule of Law (his caps) in short-circuiting the UCMJ. But he never once asks why the Corps hasn’t committed the same.

Often, the court is too reluctant to strike down legislation because it exceeds Congress’s power under the commerce clause, or because it takes private property for public use without just compensation. Justice Stevens’s dreadful misreading of “public use” in Kelo, for example, illustrates the danger of turning linguistic somersaults to allow state land planners to drive people from their homes. But neither Hamdan nor Rapanos raise any tension between democratic decision making and judicial power. Both cases only ask whether the president or the Army Corps of Engineers complied with their own statutory mandates, or chose to make new laws under the guise of interpretation.

The key element of trust should play out the same in both cases. Our Constitution starts out with a presumption of distrust of all government actors, which is why it drew a sharp line between the legislative and executive branches. We can argue until the cows come home whether national security or environmental protection presents the greater threat of executive or administrative misuse. But that ranking really doesn’t matter, because there is no reason why the Supreme Court has to defer to overaggressive public officials in either context. Justice Stevens rightly chastised the president for flouting the rule of law in Hamdan. But he was tone deaf on the easier question of statutory construction when blessing the Corps’ extravagant reading of the statute in Rapanos. We will get consistent and reliable statutory construction only when all justices put aside the naïve fantasy that they lack expertise and information to read the common language found in congressional enactments.