Olson is an economics researcher from the sub-field of "institutional economics". "Power and Prosperity" is his last book, published posthumously in the year 2000, by his friends. Most people know this book for the introduction of the catch-terms "roving bandit" and "stationary bandit".

Broadly speaking, Olson was inspired by the question "Why does not prosperity always follow a collapse of bad government?". Olson was born and spent all his life in the U.S.A., but is surprisingly perhaps better known in the former Soviet Union. Possibly because he was one of the few researchers who did not neglect Soviet type regimes when studying autocracies.

Olson is also one of the few liberal economists who do not automatically dismiss state's role in a country's economic development. He created a much less-known term "market-augmenting government", of which he was most proud of. He thus concluded that researching the structures of power is important.

As a part of this research, he examined the "succession crises" in authoritarian states and "government incentives" in democratic ones. As the simplest example, even participating in the elections requires effort on the part of the voter, which most election outcomes do not justify. (Rational ignorance.)

Therefore, the examples of reform without crises are few.

The book aims to answer the following questions:

  1. How do transitions to market-augmenting governments happen?
  2. How do implementations of reforms affect reform planning? Specifically, in what order should reforms be implemented?
  3. Are different market-augmenting institutions appropriate for different levels of development? (Also, does globalisation help?)
  4. How institutions are affected by conflict?
  5. Are grassroots (democratic?) initiatives helpful?
  6. To what extent are self-regulating institutions possible?

Remark: only recently "liberal" economists (i.e. not Marxists) started to seriously consider "political economy".

His main contribution is the study of "collective action", why it very often fails. The important terms here are "encompassing interests" and "over-encompassing interests".

One more important phrase: "Prosperity depends more on the wisdom of the population that on the freedom of the bargain."

1. TODO Body

1.2. DONE The Logic of Power

1.2.2. The "Criminal" Metaphor

  1. Humans can be different:
    1. Rational
    2. Benevolent
    3. Malevolent
    4. Stochastic
  2. A "Criminal" is someone who is acting out of pure power capacity, and is unconstrained by morals.
    1. Criminal activity, i.e. robbery, is absolutely NOT a voluntary transaction.
  3. A Criminal lives better in a rich society than in a poor society.
  4. Criminal activity decreases GDP.
    1. Therefore a rational criminal cannot steal too much, for the same reason as shearing sheep is better than slaughtering them.
    2. This logic does not actually prevent thieves from stealing.

      This is easy:

      Suppose a thief in a society of size M steals X, therefore the GDP is reduced by X. We divide X by M, \(X/M\) => this is his individual loss. His gain is X, therefore his total profit is \(X - (X/M) = X\frac{M-1}{M}\), which is very close to 1.

      We can have a more complicated formula: say, a thief steals X, but society loses Y(X). Then the criminal's net gain is \( X - Y(X)/M\). Solving the equation \( X - Y(X)/M = 0 \), we get \( Y(X) \geq M\cdot X \), which is very unlikely to happen. This is basically people provoking calamities for fun, indeed, happens very seldom.

      Sometimes the formula \(X/M\) is wrong, as the "criminal" participant has a larger stake in the GDP. This is the case Olson is interested in.

      Definition: "Stake": what one gets or gains from GDP or loses from thievery.

    3. Formulas for honest work:

      The "ex-criminal" works and gets a salary of X. His company gains \(α X\) in profits, and GDP increases by \( (1+α)X \), of which the "ex-criminal" also gets \((1+α)X/M \).

      1. There are imperfections in the market, no necessarily coming from corruption: the "ex-criminal" also needs to learn stuff, and is an agent with bounded rationality.
    4. The behaviour of a Mafia boss.
      1. A Mafia boss is different from a thief.
      2. A Mafia boss steals, but also kills other thieves.
      3. A Mafia boss therefore has an encompassing interest.
      4. The Mafia boss, therefore, does not have such a simple formula for his profit.
      5. He is typically stealing a percentage of the GDP growth per time period, and this percentage is less than 100.
      6. That is, he has an incentive to steal less when he is stealing from everyone.

      The loss to the society is \(M*\beta \), the profit to the mafia boss is \((M-1)\beta - c\), where c is the cost of his army.

1.2.3. The Stationary Bandit

  1. The story of Feng Yu-xiang defeating White Wolf. Why are stationary bandits preferred to roving bandits? Because they have encompassing interests.

    A stationary bandit actually has a higher stake than a Mafia boss, because he does not compete with the government for taxes.

    He robs at a revenue-maximising rate of theft.

    He robs for \(\beta \cdot GDP\), he loses after redistribution: \(\beta \cdot GDP/M\). Something is fishy here.

    Remark: in his article on the mathematics of losses, "The Economics of Autocracy and Majority Rule", he considers quadratic "deadweight loss" dependency on taxation.

    Suppose he taxes (robs) for \(\beta \cdot GDP\), then the actual GDP loss is \(\alpha \beta \cdot GDP \), the bandit is losing \(\alpha \beta \cdot GDP/M \). Equation: \(\beta \cdot GDP =\alpha \beta \cdot GDP/M\) => \(\alpha = M\) Does not sound too good. Only when \(\alpha\) is in millions, will the bandit stop robbing.

    1. Example

      If he cuts his rate of tax theft from 95 percent to 90 percent, he doubles his subjects’ posttax reward for production and trade, which might well increase output and tax receipts by a large multiple.

      By 95%, I suspect, they mean 95% of income, not 95% of revenue?

      (I cannot stress the importance of this.)

      Indeed, reducing 95%->90% will make peasant's obtained income twice greater.

      Assuming that the peasant keeps the same 5% for himself, and re-invests the other 5%, indeed, the money can increase greatly.

  2. Origin of Autocracy

    Civilisation also grows under autocracy, despite those leaders being absolutely selfish.

1.2.4. Historical Record

  1. The Other Invisible Hand

    (What makes the autocrat provide public good.)

  2. Princely Consumption

    What do the princes need money for? Taj Mahal, Versailles, but more importantly, "public goods on a whim" – modifying the country as he goes.

    And, of course, armies!

1.2.5. Comparing Autocrats and Majorities

Democracies are won by promising to re-distribute stuff from minorities to the majority.

The easiest arithmetic example comes from supposing that the revenue-maximizing tax rate is one- third and that the majority earns one-third of the national in- come in the marketplace. The rational autocrat will then find that the last dollar in taxes that he collects reduces the na- tional income by $3, and one-third of this loss is his loss, so he breaks even on this last dollar of tax collection and is at his revenue-maximizing rate. But if a majority mistakenly chooses this same tax rate, it hurts itself, for it loses $2 (the same dollar lost by the autocrat, plus $1 of market income) from the last dollar it collected in taxes. Thus, a majority would maximize its total income with a lower tax rate and a smaller redistribution to itself than would be chosen by an au- tocrat.

Really? Why does not the king lose the same dollar of market income?

Generally, the revenue-maximising rate is about the one which makes sure that king's share of the income falls by 1 over his tax revenue.

The point is that, in his logic, the king gets about 1/income, because taxation is his only source of income, and the majority is also getting money from the market.

1.2.6. Superencompassing Majorities

Are the majorities that do not to re-distribution.

1.3. Time, Takings, and Individual Rights

Time preference matters. An autocrat with a short lifespan always has an incentive to become a "roving bandit".

1.3.1. "Long Live the King"

Dynastic succession increases the likelihood of long-term planning.

1.3.2. How Does Banditry End?

Autocrats are usually replaced with other autocrats, or, even worse, roving bandits.

He uses Schumpeter's definition of democracy: "a system in which leadership of the present government is subject to replacement because of the free electoral competition".

Breadth of franchise is NOT the key to why autocracies are replaced with democracies.

Claim: autocracies are replaced with democracies when democracies defeat them in war (or subversion).

But why do the victors themselves not become autocrats?

  1. There are too many of them and the power is too distributed
  2. There are conditions which prevent fracturing
  3. The area is not under a threat from the outside

1.3.3. Representative Governments Institute Property and Contract Rights

Well, simply because leaders of various factions do not want their stuff confiscated.

1.3.4. Lasting Democracy Implies Lasting Property and Contract Rights

But not the other way round. And make no mistake, it's not that democracy creates property rights, it's that property rights are necessary condition.

1.3.5. lwf: summary

So, basically the second chapter is more or less meaningless. Austrian economists have shown the same results, with more confidence.

1.4. Coaseian Bargains, Transactions Costs, and Anarchy

1.4.1. Beginning with Mutually Advantageous Transactions

Adam Smith: The Invisible Hand of the Market: In an exchange, both parties profit.

1.4.2. Transaction Costs

Ronald Coase defined "market imperfections" as transaction costs.

Hierarchical firms exist as places where transaction costs are reduced to much smaller than those on a "real market".

1.4.3. Market Failures

Pollution damage, negative externalities.

Government fines must be larger than the loss of the people, to force companies to reduce pollution.

There are also positive externalities, when companies train their employees to better work, and they become better people in general.

"Pigou theorem" (1948): when externalities are present, "laissez-faire" market does not lead to Pareto-efficient outcome.

Non-purchasers cannot be excluded.

(Example: rancher's cattle is trampling a farmer's crops.)

1.4.4. The Coase Theorem

When non-purchasers can negotiate, the Pigou theorem is wrong, and through the "Coaseian bargain", a Pareto-efficient outcome can be attained, unless transaction costs are lower than the gains.

1.4.5. Generalising Coase

Can we extend Coase Theorem to non-voluntary exchange (i.e., the government)?

(My example: you give me all your money, and I do not kill you. Perfect bargain, both of us are better off: I get richer, and you stay alive.)

If we do, we produce weird results.

1.4.6. Anarchy