The Empire Bias
Firms tend to be managed with an “empire” bias, i.e., toward being too big. The largest firms are less profitable, they too often reinvest rather than pay dividends, and firm mergers and acquisitions tend to lose money. Also, as I suggested yesterday, the fact that organizations can be too big at all is probably due to a bias to over-manage top subordinates, to justify and affirm supervisor status.
The fact that we can so easily see this bias shows that investors have limited control over managers. If it were easier to buy firms out and change their management, presumably we’d see more large firms forced to break up, and more top managers forced to accept autonomous subordinates. (Which would be a good thing.) Instead, overly large firms are now disciplined mainly by having to shrink when they run out of money.
Non-business organizations also suffer this empire bias; their managers also overly focus on increasing size and overly-managing subordinates to justify and assert their status. This empire bias seems worse for states than firms because:
Financial losses quickly threaten firms, but states can be less efficient before war or revolt threatens them.
Preventing war or revolt depends less on supplying value and more on ensuring loyalty, which empire aids.
States accountable to voters or elites tilt toward their biases, and most people under-appreciate org scale disecon.
Non-pivotal voters or elites gain status by advocating more empire, even when empire policies hurt them on net.
A world government should be even more biased to overly-manage overly-large overly-intrusive agencies. After all, its existence is much less threatened by war or comparison with productive outsiders. The Soviet Union collapsed because its citizens could eventually see clearly the West’s superior productivity; a world-wide Soviet Union would avoid such embarrassment.
Of course future world governments might add great value dealing with global coordination problems. Whether a world government will be worth its empire costs depends on how serious and frequent are such problems, and on how much better we learn to structure large organizations to avoid such costs.