Economists have long debated how government spending affects an economy mired in recession. One influential view, associated with Keynesian theory, rests on what is sometimes called the paradox of thrift. According to this view, when households and firms simultaneously cut their spending in an effort to save, aggregate demand falls; because one person's spending is another person's income, the collective attempt to save can shrink total output and, paradoxically, leave the economy worse off and savings lower than before. Proponents argue that, in such a downturn, government can step in as a spender of last resort: by borrowing and spending when the private sector will not, it offsets the demand shortfall and helps restore output. Critics of stimulus advance a competing argument known as crowding out. They contend that government borrowing is not costless: to finance its spending, the government must draw on a finite pool of savings, and the increased demand for funds tends to push interest rates upward. Higher interest rates, in turn, discourage private investment, so that public spending displaces, rather than adds to, economic activity. In its strongest form, the crowding-out critique holds that a dollar of additional government spending reduces private spending by roughly a dollar, leaving total output essentially unchanged. Defenders of stimulus typically respond that the crowding-out mechanism depends on conditions that may not hold during a deep slump. When many resources sit idle and interest rates are already near zero, they argue, there is ample unused saving and little upward pressure on rates, so additional government borrowing need not displace private investment. The force of the crowding-out objection, on this account, varies with the state of the economy rather than holding uniformly.