Institutional Change

"What's the point of knowing the world if you cannot change it?"

Theories of institutional change:

• functional/efficiency theories

  North 1993: When groups see opportunities that cannot be realized under existing institutions, they make new ones. (Examples: new companies. bankers inventing new financial instruments. Property rights). North's ideas are tricky because they are so obvious that it's hard to tell if they are genius (things which are obvious to all but only after a genius figured them out) obvious or tautologically obvious: people create institutions because they need new institutions in order to get stuff they want.

• public interest
• power
• path dependence
• evolutionary
• Bayesian Learning/demonstration effect
• Crisis as trigger

Exam notes

35% of final grade. Computer lab 2. From 12 problems, choose six. For each problem, describe in one paragraph and look at appropriate institutional mechanisms. E.g., poverty, rent-seeking, regulatory capture, externalities. Think about enforcement, about principles behind mechanism. Discuss hypothetical consequences of adopting the prescribed institution: solve problem, don't solve problem, perverse incentives.

Final Paper

Due May 5.

Service Delivery for the poor

Per-capita spending on health doesn't correlate with better health conditions for poor. One key reason is that, in almost all countries, most of that spending is concentrated with the wealthy, so per-capita spending on the poor remains very low.

Property rights, contracts, transaction costs.

From an economic standpoint, legal systems' functions are to define, transfer, and protect property rights.

The role of the judiciary: they only clarify and define within the bounds of existing law, rather than creating law. Objection: in practice, this distinction is impossible; the act of clarifying often entails creating new law.

Informal enforcement mechanisms:

• Unilateral
• Bilateral
  • self-enforcing contracts
  • vertical relationship
  • hostages/collateral/third party
  • private enforcement ... violence
• Multilateral
  • reputation
  • enforcement by powerful third party (could be collective government)

The Economic Theory of Public Enforcement of Law, A. Mitchell Polinsky and Steven Shavelli

• p 3. "we assume for simplicity that public enforcement is the exclusive means of enforcement"
• p 4. "For simplicity, we focus on the assumption that individuals are risk neutral in fines and in imprisonment."
• p 19. "assuming for simplicity that injurers are always found liable"
• p 31. "assume for simplicity that individuals are risk neutral"
• p 35. "assume for simplicity that sanctions do not deter"
• In the discussion forums for my mail program (mutt), I've seen proposals for a feature whereby, if the body of the email includes the word "attachment" or "attached", but you try to send the email without any attachments, the programs asks you if you are sure. I would love that feature. On a related note, I think that all economists should be required to use a word processor which detects the phrase "for simplicity" and forces the author to include a section that addresses the probability that the simplifying assumption is true in the real world and how irrelevant the previous analysis will be if the assumption is not true.

Class discussion

A chart showing a correlation between income levels and less burdensome regulation. But the ability to go into business without any regulation presumably doesn't correlate as well, since then legitimate businesses would be competing with charlatans who are bad for the economy. How would we measure that, and try to correlate it to wealth? That is to say, a lot of legitimate businesses are quite happy to have enforced regulations, so they don't have to compete in a race to the bottom. Who would want to try to make a profit selling widgets if the competition was a steady stream of fly-by-night companies making widgets that break two days after you buy them? You would have to turn into another fly-by-night company yourself. So what is the measurement that would suss this "level and not-ground-level playing field" issue out?

A Conceptual Framework for Interpreting Recorded Human History Douglass C North, John Joseph Wallis, and Barry R. Weingast. NBER Working Paper No. 12795, December 2006

Thesis: The main social order for the last 10,000 years has been limited access order. These systems limit economic participation in order to create rents, and then using the rents to stabilize the political system and limit violence. Remember that rent in a technical economics term meaning, loosely, the excess profit you can get out of something when you don't have real competition. E.g, monopoly profits. So the paper is saying, essentially, that the ruling elites have always monopolized the economy to cheat people and then used that ill-gotten wealth to stay in power.

The rest of the thesis is that some countries have, over the last 300 years, started switching to open access order, with real competition, and that this provides a different basis for social and political stability.

Update:The paper defines three orders, primitive, limited access, and open access. I got about 30 pages in and was eagerly looking for the difference between the primitive and the limited access when I suddenly realized that, when the authors started using the term "the natural state", which I assumed was another way to say "the primitive state", they might actually mean something different, such as lumping both the primitive and limited access states together as "the natural state". Or something. Whatever "natural state" means, it's a bad term: it's ambiguous and overloaded with meaning from other disciplines. Nor is "natural" the defining characteristic of a natural state in the way that "limited access" defines, well, limited access orders. I'm still reading, but with much dismay as part of the foundation of the understanding I thought I was building is undermined.

Update 2: A classmate takes pity on my poor reading comprehension and points out where in the paper North defines "natural" as another term for "limited access". I continue to believe the paper would be better if the term "natural" was removed.

• p 5:

  Social orders are composed of constituent systems, such as the economic, political, military, and religious systems. ... Our framework acknowledges that political and economic systems are organically related, as they are both parts of the same social order.

• p 6: political and economic systems are interdependent. So "development economics and the international donor community" both keep failing to promote development because they can't or don't change both systems simultaneously.
• p 10: "every society has to develop mechanisms to ensure or restore order. The benefits of even moderate limits on violence are large enough to gain support from most non-elites as well as elites." I've thought about this in the context of China previously, but I suspect that this is the single strongest explanatory factor for popular acquiescence to dictatorship and authoritarian regimes. Note that mortality from violence is sharply up in Iraq from pre-2003 levels.
• p 11: an economic explanation of how two neighboring warlords might reach peace: the economic advantage of peace is incentive to make "credible commitment" to peace and set up a social order. Note that this mechanism only provides incentives to limit violence, not to renounce it. They're still warlords, maintaining both their internal political power and the state of peace through the credible threat of violence.
• p 12: "modeling the state as a single actor is inherently flawed. Unless we understand the dynamics of relationships within the organization of the state, we can never understand the interrelationship of politics and economics." Yay!
• p 14: "Recognizing and supporting elite organizational forms, that is by providing the institutional framework within which elites (and no one else) can form organizations, is one of the most valuable privilege that elites possess." A very formal way to say that The Man stays on top.
• p 25: "the clean break in the conceptual framework between the three orders is not matched by a neatly observable clean break in the historical record." Things are not very binary (recursion intended).
• p 26: transition from limited to open social order historically takes about 50 years or less.
• p 27: "Dictatorships, strong men, juntas, aristocracies, monarchies (hereditary and not), single party regimes, and representative assemblies (of the elites not the masses) all seem to represent viable internal structures for a natural state in some historical circumstances. ... internal revolutions ... change the faces of the leading elements in the dominant coalition without changing the nature of the social order." Q: Which order best describes Singapore? "... the same institution will operate differently in an open access order than in a limited access order. ... Elections, for example, work differently ..."
• p 30: typo: "If the coalition becomes to large,"
• p 32: the difference between the natural state and limited access orders. Er, maybe not. See my update at the top. I think they are saying that the difference between "primitive" and "limited access" is that limited access is a more impersonal, scalable version of "primitive."
• p 36. "Neo-classical economists believe that prices always work to allocate resources and have done so throughout all of human history." But this is probably wrong (paraphrasing) because elites often have monopolies but don't charge monopoly prices, as a bribe to other elites to maintain loyalty. So price-based signaling has probably been broken for much of human history.
• p 39. Broken sentence: "The ability of entrepreneurs to perceive new opportunities to capture rents and to form organizations to capture those benefits."
• p 39. The authors weigh in on the "democracy leads to economic growth" argument, by taking up the inverse position: "Sustaining competitive democracy is possible only in the presence of economic competition and the emergence of sophisticated economic organizations."
• p 39: "... the open access state suppresses alternative sources of violence. Open access orders that fail [to do so] typically fail to remain open access orders. ... Further, the illegitimate use of violence by the state can be identified and policed by sanctions against members of the government. [Footnote:] The rules about legitimate use of power must be clear, typically through constitutional “bright lines” that make it clear when those in power have violated the rule." Which would lead me to two conclusions regarding the US: first, the erosion of the rule that only the Congress can declare war is threatening our capitalism and our democracy. Second, impeaching any of the recent presidents who waged war without a declaration of war would help restore our capitalism and our democracy.
• p 43: Agreement with the "neo-Tocquevillean" view of civil society, though not using that term.
• p 45: typo: " A key way to limit the stakes of power at issue in political competition is to specifying a range of rights, including economic rights, that the government must not transgression."
• p 50: The explanation for transition must include only changes which are, at the time of adoption, in the interest of the dominant coalition. (The same principle is found in biological evolution.) So I guess radical change that ultimately harms the dominant coalition is possible only through unintended consequences, external forces, or ... what? Revolution?
• p 51: "The central feature of the transition is the development of impersonal exchange among elites." Huh. That sounds right, but is bemusing because another point that I'd been thinking about before coming to school, and which has been reinforced in classes, is the importance of interpersonal relations in successful projects and organizations. I guess that's not exactly the same scale or locus (whatever locus means) that the article is talking about, but still. Certainly my willingness to buy things over the internet, perhaps the most impersonal transaction possible, helps the economy, but I don't see that as an exchange among elites.
• p 52: the doorstep conditions: rule of law among elites, perpetual forms of organization for elites, political control of military. Perpetuality depends on the ability of office holders to bind their successors. (And hence, the Roberts Court's attack on stare decisis threatens our social order. Boy, there's a lot of strong conclusions just one step away from the text.)
• p 53: typo: "The doorstep conditions fosters impersonal elite exchange "
• p 64: typo? "When elites see this extension to their advantage, access natural states have an incentive to increase access." Another: "Moreover, having created the institutional mechanisms for maintain rule of law among elites"
• p 64: Transition mechanisms. First, fiscal mechanisms. Example: "banking. In the 1810's Massachusetts set up a system of taxing bank capital. The intent of the tax was to limit entry! But the state soon realized it obtained more revenue from the tax on bank capital than from dividends on bank stock in the banks the state was trying to protect. The state decided to sell its bank stock, and proceeded to allow open entry into banking. ... open access incorporation in banking, what is called “free banking” in the United States, began spreading widely in the 1840s."
• p 65: second mechanism: the institutions surrounding representation. Third mechanism: international competition.
• p 68: "Early corporations were created explicitly to generate rents by giving one group privileged access to an organizational form ... Once the corporation had been accepted as a legitimate organizational form, however, it was actually relatively easy to make minor changes in the institution itself that had dramatic effects on the polity and economy."
• p 69: "As a result, important pieces of the transition may occur without obvious or overt institutional change."
• p 70: "We have termed the political and economic structure of the limited access order the natural state for a reason: it is the natural form of human society." So the primitive order comes before the limited access order, and the limited access order is also called the natural state? Huh? Confuse much? And now we are changing to an "unnatural" state with liberal capitalist democracies? I think it's better just to not use the term natural state.
• p 71: "Natural states are not failed states, they are typically not produced by evil men with evil intentions, and they are not the result of pathologies in the structure of these societies. Nothing is unnatural about natural states. And because natural states are not sick, policy medicine will not cure them." But they are the result of pathologies in the structure of these societies. Just like the tragedy of the commons, individual people act rationally but the society is irrational. Q: What is that if not a pathology? Perhaps they mean that the culture, rather than the society, is not pathological?
• p 73: "only eight countries have made this transformation [to an open order society] since WWII. [footnote:] Moreover, none of these countries were the focus of international donor agencies." Q: Which eight? Korea, Taiwan? Some in Europe?

What are the fundamental problems of political institutions?

• How does government control the governed, and how does it control itself?
• How do we reveal and aggregate preferences?
• How can we provide public goods?
• who will guard the guardians
• Joel's note: how about, how to guarantee the rights of the minority?

Federalism and compound republics

• Individuals, not the state or the collective or the church, are the basic unit of design and analysis. C.f. China or Singapore: the interest of the state trumps the interest of individuals
• government's purpose is to use individual interests to constrain and order individual behavior. "that the private interest of every individual may be a sentinel over the public rights." (Federalist) C.f. Marxist theory, that "class consciousness" will motivate people. Joel's note: I guess, putting it that way, that both are right. People do act out of individual interest, but that interest is determined in a cultural context in which class is very important
• When individual, rational pursuit of self-interest results in an emergent irrational result, we have a collective action problem. Republics solve this by providing a mechanism to reveal and pursue a more rational collective outcome.
• Individuals can learn, to institutional design is always experimental.
• covenant approach: do unto others what you would have done to you.
• Government institutions are pluralistic. There are 89,527 units of government in the US. (About 3000 counties, 36,000 cities and towns, and 50,000 special districts, half of which are school districts.) C.f. Indonesia or the Philippines, where most provinces have only limited authority to create governments (although there are autonomous exceptions: Aceh, Kashmir, etc).
• The constitution provides the basic rules for everything else, including rule-making.
• Perpetual contestation is built into the system. Conflict elucidates information, clarifies alternatives, stimulates innovation, helps achieve complementarity of interests. Stalemates are better than a dominant center.

Besides executive, legislative, and judicial, what other institutions often serve as branches of government? The military, media, civil society, the monarchy (yes, in some countries the monarchy still plays a meaningful role), the church.

Discussion: When you have a democracy, it's unlikely to go back to authoritarianism. But Pakistan has had a real democracy but had the legitimate president suspend the elected national government twice and local government once. Both the president and the government were elected; who should trump whom?

Ulysses as an example of the credible commitment problem. In order to prove his promise, he ties himself up. Joel's note: bad example: Ulysses ties himself up because he needs to be able to hear the sirens without acting under their control. He needs to be able to hear them so that his crew, all wearing earplugs, know when it's safe to take out their earplugs. Which raises the question of why the sirens didn't force Ulysses to prematurely signal the end of danger. I guess their control wasn't that granular. But either way, Ulysses and the sirens was just a tactical problem, not a classic commitment problem.

Which elements of federalism preserve markets in the US?

• Single-party system provides a credible promise that government policy will not change over the near or medium term.
• Hierarchy. At least two autonomous levels controlling the same territory.
• Trade is protected as the federal level (an example of both the minority (free traders) having protected rights over the majority (people who would gain from parochial barriers), and of the collective action problem being solved by ceding local power to an aggregated authority).
• Hard budget constraints on lower levels. Only the national government has any legitimate reason to run deficits.
• Restrictions on federal government plus competition between lower levels limit the growth of economic regulation.

How does federalism interact with markets in China?

• Hierarchy. Because of the multi-level hierarchy, there is a series of principal-agent problems.
• sub-national governments have regulatory control
• Dependence on foreign direct investment
• Hard budget constraints on lower levels.
• Corporations are the main contributors for local government revenues.
• State-owned enterprises pay the government, which redistributes back to local governments. Local governments provide most of the social goods.

  Discussion: 18 million people in Shanghai. Medical care comes from employers; unemployed people do not have substantial safety nets. Gamesmanship between provinces and central government. 75% of Chinese are farmers, and when they are cut off from their land, they go through the cracks. New initiatives to organize groups of farmers into shared medical and pension funds.

• Individuals pay income tax to the provincial government, which forwards it to the central government, which redistributes among provinces.
• Provinces compete for big projects funded by the central government.

Problems with decentralization

• Diffusion of authority, leading to bargaining between central and local governments
• Unequal distribution between local units
• If local governments can borrow, the total government borrowing at all levels could be excessive and come at the expense of private investment. Happening now in the Philippines.

The case of Singapore

The framers of the US constitution did not trust single parties; Singapore's philosophy is that the single party is suited to run the country, and depends on internal controls for checks and balances. In China, smart people join the party to get ahead. Singapore recruits the smart people. In Bangladesh you join the army to get ahead and later go to politics. In other countries families determine opportunity and succession.

Side discussion, eventually reaching the point of emigration. Here's the punchline:

Consider this: every year, 6,000 to 7,000 Singaporeans leave to settle down overseas, including many professionals. This is 15% of today's annual births, probably the highest proportion in the world.

One website survey, which is unverified, has put Singapore's average outflow at 26.11 migrants per 1,000 citizens, the second highest in the world - next only to East Timor (51.07).

Mechanisms for preference revelation

• direct democracy
• representation

LKY prefers a trustee model to a representative model; if politicians represent narrow constituencies, they will promote factions and racial disharmony.

Arrow's theorem: it's impossible to have voting which is both fair and consistent. Joel's note: yes, but only if there are more than two choices. Bottom line: those who control the agenda can manipulate social choice.

Direct democracy problems: tyranny of the majority. Bundling problem: candidates come with an array of positions, and you may agree only with some of them but you're stuck with all of them.

Other mechanisms: Singapore. Government determines preferences from feedback units, meet the people sessions, people's associations, appointed opposition. Narrow interest groups can dominate diffuse public interest.

Principal-agent problem where representatives spend public resources to win elections.

How political contestation affects the economy

Proportional rule leads to more parties than majority rule; more parties leads to more pandering and higher budget deficits. Example: Indonesia vs UK. Also fiscal reform is harder.

Problems with representative government

• provide an avenue for special interests and geographic constituencies to take from the whole
• electoral cycles and the need to address constituents distort policy

Discussion: it doesn't seem like the representative system, and the US constitution, really address the economic issues. It was designed for political reasons, not economic considerations.

Joel's Q: Modern political discussion, especially since WWII and even more especially since 1989, is completely dominated by economics and, in particular, with growth and "developed vs developing". What was it like hundreds of years ago?

Modern institutional design in liberal democracies includes divesting some fiscal powers to an independent or semi-independent entity.

Adam Przeworski; Fernando Limongi, Political Regimes and Economic Growth, The Journal of Economic Perspectives, Vol. 7, No. 3. (Summer, 1993), pp. 51-69.

This is of course one of the $64,000 questions in modern political science: what is the relationship between democracy and economic growth? And do China and Singapore (and the other Asian tigers) prove the latter doesn't depend on the former? Let's see what they have to say.

• p 53: "The property rights literature treats the state as the only source of potential threat. But property rights are threatened by private actors: capitalist property is threatened by organized workers, landlords' property by landless peasants." Um. Talk about missing the point. Yes, it's true that the state is not the only threat to property rights, but it's a bit blind to cite workers and peasants as the next threat. Not to put too fine a point on it, how did the landlord get that property in the first place, and why are the peasants landless?
• p 54: Democracy can't be taken for granted as a defender of property because the median person will have more power as a citizen then property as a market participant, so democracy will always be prone to redistribute property.
• p 56: There is an argument that authoritarian governments can pursue autonomous governance (meaning free from the influence of, in particular, unions and large firms) that uses the best possible policies to get growth.
• p 56: So far, this entire paper reads to me like a parody of economists who seek only abstract economic explanations for the world, completely ignoring human behavior or relegating it to footnotes: "Dore (1978) offered a culturalist explanation: 'I suspect that a major motive [of dictators]. . . is to increase national 'strength' and prestige ... and thereby their own position in the ranks of the world's rulers.'"
• p 57: More reasoning on why authoritarian states might be more autonomous, and therefore potentially better at running economies. Finally, the word "behave" appears: "What [this reasoning] fails to answer is why an autonomous state would behave in the interests, long- or short-term ones, of anyone else."
• p 58: Thinking in terms of who gets to decide how big the government is and who gets the excess economic output. Democracy, citizens decide both; autocracy, state decides both; bureaucracy, state decides how big it will be but citizens get the "fiscal residuum".
• p 60: we conclude an un-edifying tour through the literature and move on to a survey of the data. But "what is even more puzzling is that among the 11 results published before 1988, eight found that authoritarian regimes grew faster, while none of the nine results published after 1987 supported this finding. ... one can only wonder about the relation between statistics and ideology. For reasons discussed below, we hesitate to attach much significance to these results one way or another. Hence, we still do not know what the facts are." So the point of your twenty-six page paper is to tell us that neither the theoretical literature nor the existing statistical experiments can tell us anything about the relationship between type of government and economic performance? Well, a null result is information too. But why should I read the rest of the paper?

As an aside, note that an unspoken and fundamental assumption of the paper is that GDP growth is the sole important measurement of economic welfare. This is so pervasive that the authors write that "Observing Brazil in 1988, we discover that it was a democracy which declined at the rate of 2.06 percent." Brazil's economy didn't shrink 2.06 percent, Brazilians didn't become 2.06 percent poorer or 2.06 percent less happy: Brazil as a nation and as a democracy declined 2.06 percent. Ouch.

Thomas Mann is a guest speaker in class today, and we have an extra dozen Korean visitors from Jungwon in class, doubling class size. Mann:

The election ... but whoever wins, they'll discover all of the limitations of the institution of the presidency. Campaigning and governing are different tasks, but they have effectively merged in US politics. Campaigns last for years and governing involves using campaign strategies to build support. ... However, there was no golden age in American democracy; the institutions have never performed exactly as intended by the framers.

Two important properties of American governing institutions: separation of powers and ...

The first article of the constitution is about Congress, not the President.

Political parties are not mentioned in the Constitution, and the framers were very worried about parties.

There has been a natural accumulation of power to the president under conditions of threat or perceived threat. It will be interesting to see the extent of the backlash and whether the next president will be similarly inclined.

Q: in many countries, the armed forces constitute a fourth branch of government. How has the US avoided this problem? A: Civilian control of the military has always been a central idea. Some generals have gone on to the presidency, but civilian control is a strong norm. (Note that Michael Hayden is serving as director of the CIA while still holding an active miitary commission. Johnson and Carter also appointed active military to this job, so it's not just another Bush attack on our norms)

Two biggest perceived problems with institutional design of US government: electoral college, and under-representation of big states in Senate.

Q: In a system where lobbying is allowed, how do you balance the difference in lobbying power between businesses and civil society? A: We haven't solved this in the US.

Class lecture

Back to property rights, public goods, etc. Property rights in extreme:

Disclaimer: pressed for time, summaries from skimming only.

Rubin, P. (2005). Legal Systems as Frameworks for Market Exchanges. In Menard C. and Shirley, M.. Handbook of Institutional Economics. Elsevier.

• Legal institutions last for centuries. Countries' legal systems very clearly reflect who colonized them, and even vary depending on what fraction of population the colonial settlers actually comprised.
• You can't just graft on a foreign legal system.
• legal protection of private property correlates well with economic growth. To put on the Marxist hat for a moment, that's not exactly a shocking surprise, especially since the term "economic growth", as used, implies a lot about the nature of the economy being measured. In other words, protecting property rights correlates well with the growth of a system in which some people own a lot of stuff. In that light, it's a bit of a "duh?". From there we circle back to and perhaps dead-end at the usual arguments about GDP versus other measures, productivity versus happiness as the ultimate goal, self-determination, paternalism, etc. Just a thought.
• "The major cost of opportunism [cheating] is neither the cost of cheating, nor even the cost of precautions taken .... Rather, it is the lost social value from the otherwise profitable deals that do not transpire." This sounds dead-on. China's miracle is in getting everybody to work in productive ways. In broken economies, wealth creation just doesn't happen much; people probably still expend a lot of effort, but not on creative activities.
• Common law evolved out of competing legal systems in the same country and embodies the victory of Merchant Law over the King's Bench, the Exchequer, the Court of Common Pleas, and other suppliers of judiciousness.
• Implications: institutions are hard to change. Rule of law and property rights increase economic well-being. Competing legal systems within the same jurisdiction are useful for generating efficient outcomes.

Posner, R. (1998). Creating a legal framework for economic development. The WB Research Observer Vol. 13 No. Feb. 31998. p. 1-11

When trying to improve the legal system in a developing country, as seems to be required or at least helpful for growth, it may make sense to focus on getting precise rules first, and adding staff and other institutional improvements later. Copying legal rules in bulk from other countries, with some adjustments, might actually work fine.

To the extent that the business community in a poor country has its own law, it may be better to codify that law than to try to borrow another country's model. But the law may be underdeveloped in a poor country ... and the task of codification may require technical skills of drafting and organization that are in short supply. ... It is important, however, to adopt the imported code to the local culture, a task for local, not foreign, lawyers ... I do not advise dispatching European or American lawyers to tell acountry how to adapt foreign laws to its legal and social institutions and stage of economic development.

Meanwhile, don't completely neglect the judges. One cost-effective idea is to offer very big pensions, which judges forfeit if removed from office for corruption.

As for granting extensive rights to criminals, this is bound to undermine the efficacy of the criminal laws, and by doing so, unsettle property rights. Rights make it harder to convict the guilty as well as the innocent.

I am so not happy with the idea that Posner is a lifetime-tenured sitting judge in my own country. Let's at least hope that, given the chance, he would rewrite that sentence to say "As for granting extensive rights to the accused ...". He's already decided that property rights are more important than civil or human rights (a possibly unfair simplification, and I'm running low on time again so I'll just note in passing that this is back once again to the "does capitalism need democracy, Asian Values, Singapore/China as example" question), the least he could do is concede that not all suspects brought before him are necessarily guilty.

Judicial Institutions, World Development Report (2002). Chapter 6.

• Developed countries have about one professional judge and six judicial staff for every 10,000 people
• Key themes in judicial reform: increased accountability of judges, simplification, increased resources.

Note, by the way, a still-unfolding judicial corruption scandal in Malaysia. Note further that Anwar Ibrahim leaked the tape that started it all. Another data point supporting the notion that the main way in which the shell around the corrupt elites cracks is from within?

Djankov, et al. (2002). Courts. (Manuscript)

Based on studying the exact rules for evicting tenants and collecting on bounced checks in 109 countries, the authors conclude that the higher formalism of civil law (what the French have) doesn't work as well as common law (what the US and UK have), which may have hurt developing countries who got a civil law transplant.

Dam, K. (2006). Legal Institutions, Legal Origins and Governance. Chicago Working Paper Series.

What is institutional analysis?

The relationship between rules and behavior (you can think of it as law and economics (Joel's Note: I think I'd rather think primarily in terms of sociology and psychology, and I'm not the only one: "The opening chapter by Bob Goodin attempts to 'name the parts' by identifying the social science literature relevant to institutional design and formulating normative design principles. But it is not clear that Goodin's list is comprehensive (it excludes anthropology, social psychology, cybernetic control theory)..."¹)).

Institutional design is a subset of institutional policy.

Prof brings up US primary elections as example of diverse institutional design. Classmate question about why there is no secret ballot, so as the American I explain a bit about the primaries. What's interesting for this class is that the US primaries are a nearly perfect experiment into institutional design. Obama is winning most of the caucuses; Clinton most of the primaries. Is this a consequence of the different institutional designs? What are the pros and cons of winner-take-all (the predominant Republican design) versus proportional allocation of delegates? The downside to analyzing the primaries, though, is that the ultimate goal, producing a winning candidate, is much harder to measure since there are too many variables and not enough general elections to really tell what happens.

Two views of Institutional Analysis and design: institutions can be analyzed with economic tools; or, we aren't able to design good institutions, only markets can do that (Hayek's point is actually more subtle: just as he thinks markets produce collective wisdom that no one person can match, he thinks that collective tradition encodes more wisdom than one single person can match. Sounds interesting and, in the quick paraphrase I just put together from skimming a few secondary sources in class, wrong; will have to dig in more later to see what he actually thought.)

Frameworks

SSP. IAD. (Hmm - the week 3 readings)

IAD: institutional analysis and development. The "rules of the game": the properties of the goods (selling water is different than selling electricity), the community (the PAP in Singapore can recruit all of the good politicians and leave none for the opposition; this is impossible in India), the institutions. From this context, the unit of analysis is the transaction. Examine patterns of transactions within the incentive structure, and what outcomes result, and how those outcomes feed back to the context.

Joel's note: this is consistent with what I saw in Denmark: having a small, homogenous population makes certain institutional designs possible and makes other capabilities unnecessary. At least, until a bunch of outlanders immigrate and then the society stutters a bit because heterogeneity breaks one of the basic design assumptions of the system. Some class and race issues in the US can also be understood in these terms; a good-old-boy network is an institution; job interviews are an institution; the steps necessary to identify an opening and get an interview are an institution.

Two-thousand year old rice terraces in mountains in the Northern Philippines. Why have these terraces persisted and been maintained? What institutions have made this possible? Water supply, water management conflicts between the top and bottom, construction, rebuilding after storms or earthquakes, etc.

Rules

These are the control knobs for institutional design.

Boundary rule. What constraints keep players in or out? In markets? In politics? In games? E.g., in the Philippines there is no divorce: no exit rules for marriages. In Singapore, bankruptcy as a tool to keep people out of politics. Saudi Arabia: no exit rules for the monarchy. Property rights (I don't see at first glance how property rights relate to boundary rules) The fact that it's technically impractical to build mines on Mars is a de facto boundary rule; the international agreements and norms to not exploit Antarctica are de jure rules; they are both examples of boundary rules.

Position and authority rules. Even in the Philippines terraces case, there is a ranking, from the high priest and low priest on down. Buyers, sellers, brokers.

Aggregation rule: how preferences are aggregated. (That's a misleading name, if this is really about preferences.) In markets, price aggregates everything. You can vote with your wallet. Voting rules ...

Scope rule

Information rules: transparency, disclosure. Formal laws. Informal norms and customs (in some cultures, income is a private subject; many civil services disclose everybody's salary).

Payoff rules. Winner take all elections. That is, democracies where the winning party can dispense spoils and cronyism. Similarly, unstable systems subject to coup are stuck in a feedback loop, because it's hard to take power militarily, killing people and making enemies in the process, and then feel personally safe giving it up to a democracy.

Behavioral assumptions

Homo Economicus assumptions: unlimited information processing power, consistency, utility-maximizing.

Behavioral science assumptions: bounded rationality, emotions, behavioral regularities, capacity for learning. Shame in Western cultures and Eastern. Classmate mutter: Monica ... it would be very different in our country. China is willing to pay a billion dollars (more!) for pride.

1. "The Theory of Institutional Design", Robert Goodin, Review author[s]: Christopher Hood, Journal of Public Policy, Vol. 16, No. 2. (May - Aug., 1996), pp. 231-232.

Update: Another example of how institutional design affects behavior. The institution here is the ordering of author names in published research.

I have participated in both alphabetically ordered and non-ordered authorship cultures and think that there are problems with both schemes. Alphabetical ordering seems simple enough, but if anyone with a surname after D is going to be disadvantaged in their career, then that system is not as objective as it might seem.

Weimer and Vining (1998 or 2004). Policy Analysis: Concepts and Practice. Prentice Hall. New Jersey. p. 159-195

World Development Report (2002): Chapter 5: Government

Despite their reputation as rigidly orthodox "Washington Consensus" institutions, a lot of IMF and World Bank papers are startlingly non-dogmatic and useful. This one, though, seemed to have a bit of a sneer lingering between the lines.

Underdevelopment

(Reading notes. One of the key books is available from Amazon as a Kindle e-book, but this won't help our blind classmate because the Kindle format doesn't seem to be compatible with screen readers. On a side note, I found this very cool Firefox plugin, Fangs to help sighted developers see (rather than hear) what their pages look like to blind people. If you try it, you have to read the FAQ because the installer has bugs.)

"Why are a few countries rich while many are poor?" You can ask William Easterly, Gundar Midal (?) ...

Some common explanations are: under-investment, lack of technological innovation, lack of education, geography, the resource curse, poor macroeconomic policies. Next we'll walk through data contradicting all of these explanations.

Easterly's book is the leading edge of non-institutional explanations. Harrod-Domar model: growth depends on labor and capital. But the data doesn't show a relationship between investment and growth.

The poorest three fifths of countries have had zero growth of income per person since 1980. Nigeria and Hong Kong both increased capital stock by 250% from 1960 to 1985, but Hong Kong is first-world and Nigeria is not. Similarly, Gambia and Japan both grew their capital stock 500%. In both cases only country's output per worker actually grew. My Q: foreign oil companies presumably own all of that Nigerian capital stock, so would it be more helpful to look at growth in domestically owned capital stock? Classmate response: not exactly who owns it, because these statistics may already take that into account, but who has access to it.

Growth in schooling doesn't correlate to GDP growth.

Geographic explanations. But why would Singapore succeed while Malacca or Bintan doesn't?

The prevalence of dangerous diseases. Also the effects of malnutrition, especially malnutrition, in making populations stupid. Australia is reaching water-based population limits, which may be as low as 12 million depending on how much they distort their economy with water-wasteful agricultural subsidies.

Argentina was one of the five richest OECD countries in the 1920s. The Philippines was second only to Japan in GDP growth rates in the 1950s.

Lee Kuan Yew said Singapore grew faster because it couldn't waste any time considering a closed economy, and because it had no wars or serious internal conflicts

Dependency theory: former colonies continue to have disadvantaged economic links with their ex-colonists, and this drags them down generally.

No single factor is a convincing, statistically strong explanation.

institutionalist explanation

Weak, missing, or perverse institutions are at the root of underdevelopment. Many theories of development come and go like fads or fashions: protestant vs catholic work ethic, post-colonialism, etc. The current fashion is institutions.

Joel's note: from the slide: "Which institutions are needed for development?" This relates to Non-State actors class, and the "Neo-Tocquevillean" view that healthy civil society is part of the underpinning of successful governance. Of course, governance != development. But civil society is very close to "social capital", which is a nice box on the current slide, "the Evolution of Development Thought."

Old joke: Suharto visited Marcos in the Philippines. They looked at a bridge, and Suharto said, the bridge isn't finished. It stops near the end. Marcos tapped his chest and said, ten percent. Later, Marcos went to visit Suharto. Suharto showed him a view and said, you see that bridge over there? Marcos said, I don't see any bridge. Suharto tapped his chest and said, one hundred percent. (more like this)

Chicken and egg: are you rich because you have strong institutions, or do you have strong institutions because you are rich? Dani Rodrik tries to address this with econometric analysis.

"In the US the constitution is a living document." Singapore inherited working British institutions, especially compared to Bintan and other potential rival cities, because it was the hub of the British empire in Southeast Asia.

Two kinds of institutions needed for development: First, those that encourage trade by promoting trust and lowering transaction costs. Contracts, enforcement, etc. To do business in Japan, first you socialize. In Chinese, you build guanxi. (Note in both cases the importance of taking drugs together in public as a trust-building exercise.) Second, property rights: separation of powers (which China doesn't have, so how do they protect property rights well enough to have a functioning economy? Note that private ownership of land is still illegal in China, so alternatives have been devised.), keeping the government helping instead of stealing.

Joel's note: side research prompted by discussion during break: I had no idea there were as many as 13 Jewish Senators in the US. Arlen Specter is a Jew? Was I supposed to know that?

Institution explanations for underdevelopment, based on research "not storytelling"

• colonial heritage. Spanish inheritance: a small number of rich families, powerful church. French: rigid bureaucracy. British: capitalism (but c.f. India). Classmate point: Greenspan argues that Fabian socialism influenced the first-generation leaders in India. British brought decentralized government to India. Common law heritage protects property rights and maintains independence better than civil law. What about Japan? It wasn't a colonizer long enough to really change institutions. Classmate: In Bangladesh we say "heritage" for positive things and "legacy" for negative things.
• colonial heritage plus. Colonizers designed institutions to exploit locals. The Eiffel tower was built with African iron. Extractive institutions. Philippines and Latin America: Catholic church historically used to keep the population controlled, not educated. By contrast, Methodist schools in Singapore are a primary source of the educated elite in Singapore. The French didn't have religion—they colonized for wealth, power, etc. Joel's note: that's why everybody colonized; some justified it with religious missions. The global sugar market still reflects colonial heritage. Slavery.
• Political conflict. Not enough political competition, so rulers were able to build institutions to serve selfish interests. Cold war led to support of bad leaders who otherwise might not have prevailed. CIA is associated with coups. Great quote: "The only mistake Solo did was to try and stage a coup while drunk. They started playing reggae on state radio and TV until soldiers loyal to the government regrouped and overran them."
• Beliefs and norms: culture is inhospitable to markets or trust. Rich Middle east countries may seem like a counter-factual, but that could be transient, based on oil money.

These explanations are not sufficient; they aren't adequately supported by data, by econometric experiments.

One school of thought: countries succeed when institutions constrain government to keep it honest. The Suharto myth is developing as, he did good things for Indonesia at first but after he'd been in charge for a while he got too corrupt. That's especially a feedback property of any political system where violence is required to take power: once you have power, you can't retire safely, so you have no incentive do anything but build military power and live it up.

Empirical studies of which institutions matter for development. The following are significant factors in regression studies:

• protection of property rights, contracts
• civil liberties
• political rights and democracy
• political instability
• cooperation-promoting institutions.

How can countries change their institutions?

Changes are usually small and incremental. (Except Thailand, where they do things like this: "young King Ananda Mahidol returned to Thailand for the first time in 7 years. ... A new constitution, Thailand’s most democratic to date, was drafted in his honor.")

Imported institutions often fail.

Institutions are embedded in peoples' minds, and so don't change easily. Forces resist changing institutions. Fundamental institutions are long-lived and path-dependent.

Aid doesn't change norms and beliefs; Aid agencies often promote Western best practices rather than local experimentation and adaption. Short projects, rapid turnover, staff rewarded for work, not outcome.

Aid does cause some changes: privatization, changes to laws and regulations. World Bank cuts and pastes contract and plan text between different projects in different countries.

Joel's note: apparently we won't be talking about the Schmid reading today.

Rodrik, D.2000. Development Lessons for Asia. Asian Development Review. Vol. 23. No. 1 p. 1-15 (PDF) (Browse, get the substance)

Schmid, A. 2004. Conflict and Cooperation: Institutional and Behavioral Economics. MA, Blackwell Publishing. Chap 1-4 and 6-7 (MUST READ)

• p 1: "A constraint on one person is an opportunity for another." Not necessarily. Maybe everybody is constrained.
• p 2: "This book intends to form a new synthesis and assembly of ideas from many writers who do not necessarily consider themselves institutionalists. It is time for institutional economics to emerge from its critical role." That sounds good to me. But I'm disappointed that author even feels the need to mention that he's plucking ideas from people who don't self-identify in a particular -ism.
• p 4: "Chapters 3 and 4 lay a foundation for the analysis rooted in cognitive science. Bounded rationality and the limited information capacity of the human brain are fundamental." Yay! A theory about how large groups of humans interact which uses humans, instead of rational actors, as its foundation.
• p 7: I like the shorter sentence lengths, which make the text less brutal to read than, say Finnemore. But this is a bit too short: "Language is the carrier of formal rights. Sharing a language facilitates the understanding of relative opportunities. The meaning we attach to words is vital."
• p 7: The proclivity to stand patiently in line instead of jumping the queue is a piece of social capital.
• p 12: impact analysis: if the institutional alternative was different, what would be changed in terms of who gets what.
• p 13: change analysis: if the institution was different, .... What's the difference between changing the institution and changing the institutional alternative? Changing the institution seems to mean changing the nature of the institution, not changing to a different institution; perhaps changing the institutional alternative means changing the decision the institution made?
• p 17: "One of the important impacts of institutions is how they affect how the costs and benefits of inevitable surprises are shared." bingo. imagine the institution to be a software development contract, and compare fixed-price to time and materials.

  People have administrative transactions (orders), bargained transactions, customary transactions, and threat transactions.

• p 20: the staccato tumble of sentences that each have a fairly deep meaning but are packed into too few words, and not strung together well or at all, is starting to get to me. "A view of behavioral regularities from behavioral science can be used in both impact and change analysis. Counterfactual assumptions work on as "as if" basis only in limited instances." This is English? This is madnessjargon.
• p 23: "Institutional economics is not concerned with changes at the margin, but rather instances where the change in a variable is large enough to feedback on other variables." 1) most economists seem to worship the very concept of the margin, so somebody is really wrong here. 2) feedback is a noun, not a verb.
• p 25: "It is a tragedy when individuals make their most advantageous choice and wind up where they do not want to be because of the emergent aggregate effect of others doing the same thing. This is not the tragedy of the commons but the tragedy of isolated individualism. " Huh? That's the tragedy of the commons.
• p 61: "The previous chapter focused on how the individual brain works. But, to understand behavior, we must examine systems of brains — the brain in a social setting."

  

• p 68: This guy is getting pretty frustrating, because he's clearly on the right track, but it's maddeningly hard to actually learn from his sequence of little pebbles of ideas. Here is a sample from the conclusion to the chapter about systems of brains:

  People's learning is shaped by institutions, and their learning shapes evolving institutions ... institutions, both informal and formal, are repositories of knowledge.

• p 91: "When most people think of property rights, they are thinking of these incompatible use goods (IUGs)." I suppose those are "rival" goods? Now we have three terrible terms for the same thing, "rival", "incompatible use good", and "IUG".
• p 92: "Pareto-worse changes in rights are essential for some varieties of economic development." Yay! The interests of the people who happen to be rich at this moment are not absolutely inviolate. (somebody tell the RIAA)
• p 92: instead of viewing air pollution as a by-product of steel production, view it as a necessary input. That simultaneously makes a lot of sense and seems silly.
• p 99: Eight ways to unseat the free rider. A few: Rules. Trust (e.g., play more than once and use tit-for-tat). Play tit-for-tat (I just said that!). Create environments in which people learn to care for each other (social capital). Okay, he's going through the two-by-two matrix for public goods. Do I have to read it all to find out if there are any new ideas?
• p 126: "Fundamental uncertainty means 'Choice is in the first place ... the creation of choosables'". That sounds about right. Let me suggest a sibling koan: Making decisions is in the first place recognizing decisions.

p 147: results of economics experiments in the lab. "hard-nosed game theory cannot explain the data. ... It is possible to provide an environment in which at least 90 percent of subjects will become selfish Nash players. It is possible to provide an environment in which at least 90 percent of subjects contribute toward the group interest. Why and how often this all works remains a mystery.

Ostrom, E. (2005). Doing Institutional Analysis: Digging Deeper than Markets and Hierarchies. Chap 30. In Menard and Shirley (Ed). Handbook of Institutional Economics. Netherlands. Springer. (browse, get substance)

Ronald Coase sounds like he may be a big theoretical pillar for this class. Also Elinor Ostrom.

(note to self: look into Kelso.)

I don't know why, but all of my electives have around twenty or fewer people, compared to last semester when they were all thirty or more. This is much better.

Student introductions. My own interest: why people usually fail to function in groups. It's formally called the "collective action problem". Other classmates: what can we do with institutions to fight corruption in the Phillipines? Polish management consultant: Asian institutions. Bangladeshi: how to design more efficient institutions? (response: one idea is to look at multi-nationals doing business in Bangladesh as a source of (better) norms.) Second Bangladeshi: the judiciary has just been separated from the executive branch and so everything is changing. Improvement of institutions to deliver services. Institutions and development.

What is an institution?

6. a. An established law, custom, usage, practice, organization, or other element in the political or social life of a people; a regulative principle or convention subservient to the needs of an organized community or the general ends of civilization. —OED

Joel's note: OED definition 7 is "An establishment, organization, or association, instituted for the promotion of some object, esp. one of public or general utility, religious, charitable, educational" but I think that meaning is secondary for our purposes.

Humanly devised constraints that structure human interaction ...—Douglass North

Transaction costs. Joel's note: "A more recent study in the International Journal of Health Services found that in 2003, administration costs in the US health care system ate up about $400bn, or about 25% of total health care spending. (source)

Examples of institutions as constraints. Omertà. Singapore's official secrets act. Out-of-bounds markers. The Ten Commandments. What makes an institution formal? It's sanctioned by an authority. (Joel's note: and documented? but surely pre-literates could have formal institutions?) Institutions can increase or decrease transaction costs.

Ostrom's definition: "Shared concepts used by humans in repetitive situations ...". Rules: mutually understood and predictably enforced by agents. Norms: shared prescriptions enforced by actors. Strategies: regularized plans that individuals make within the structure of incentives produced by rules, norms, and expectations of others' behavior. Football is an example of a zero-sum game. Joel's note: but professional football (US or world) is actually a positive-sum economic activity; the teams cooperate to build a bigger business. Even in amateur sports, pleasure and safety are often more important than winning. I hope it will always be this easy to turn zero-sum activities into positive-sum.

J. R. Commons: bundles of rights, duties, liberties, and exposures. "...An institution is defined as collective action in control, liberation and expansion of individual action." Joel's note: The slide says 1968, but apparently he died in 1945. Which helps explain why he never won the (Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred) Nobel, which is only awarded to living people and was first awarded in 1969.

Game theory

classmate question: where is legitimacy in these definitions? Why is the context of legitimacy important in defining institutions? It's helpful in analysis, but I don't see that it's needed in the definition.

Driving patterns as national or local institutions.

Game theory: "institutions are equilibria in repeated games." Joel's note: the word "culture" is mentioned for the first time in class.

Sports analogy: institutions are de facto rules of game; law is formal rules backed by sanctions; organizations are players; governance is the actual play; public policies are outcomes. Joel's note: why aren't outcomes outcomes? I guess policies are outcomes, if we mean de facto policies, not nominal policies. Marijuana use in the US is a good example of broad and varied divergence between de jure and de facto policies.

Tragedy of the commons. Joel's note: A very tantalizing note in that paper (which defined the term): "However, given an infinite source of energy, population growth still produces an inescapable problem. The problem of the acquisition of energy is replaced by the problem of its dissipation, as J. H. Fremlin has so wittily shown. [4]" so I'm going to have track down that reference. Fortunately for J. R. Commons, the paper is from 1968, so he narrowly avoided a lifetime of terrible puns.

What is an organization?

• strategy
• structure
• systems
• skills
• style
• staff
• shared values

Joel's note: one of those things is not like the others. Organizations are collections of people. Everything else is dependent on that, different abstract concepts to better understand how these people are behaving.

Shared values: the job of the leader is to define and shape the shared values. All organizations have leaders. "You may call it first among equals but still he's first." Joel's note: how strong is the scientific consensus behind that statement? Do groups of humans inevitably have leaders?

The value of constitutions. Related concept: Stare decisis. A classmate provides an example from Bangladesh, where a general wanted to change the allocation of court benches in different districts, but the Supreme Court was able to prevent this violation of the constitution. Of course there's been a big constitutional crisis in Pakistan recently (though hardly for the first time), and in Turkey.

Informal institutions change more slowly than formal institutions. Examples of enduring informal institutions: suppressed religions, like Catholicism in China. Islamic rules in Saudi Arabia. Some of that has been formalized.

Classmate note: US constitution is stable because it's made very hard to change. Why? One reason could be that any change affects some vested interests. Joel's note: I hadn't thought of it from that angle before; I only thought about majority/minority issues. How does the government constrain itself? Actually, I guess the answer combines both issues: if you want to create a stable institution, design it so that self-interest keeps it stable. And also path-dependence: other countries could have the same rules in their constitution, but the US history has created a large, complex culture which adds its own inertia.

Role of institutions in markets

Example: standards. 16th century rice markets were dominated by big private traders, because only big players could absorb the high risks of shipping in feudal, pre-industrial Japan. New technology in the form of improved transport destabilized the market by allowing small traders to compete. This led to demand for standards, which in turn enabled more trade and new institutions (rice exchange notes, inventory credit, loan collateral, credit). (Joel's research note: Japan's first railroad opened in 1872)

Second example: bankruptcy rules. In the UK, bankruptcy rules evolved out of ad hoc arrangements between borrowers and lenders, leading to formal commercial laws to the point where a couseling company writes, "Bankruptcy in the UK is a humiliating and public experience, it is also a very primitive and grossly unfair means of dealing with a debt situation, it is a no-win situation for all involved. In the US, Congress intervened many times in the development of bankruptcy norms, leading to laws that are more debtor-friendly and have the primary goal of keeping bankrupt companies functioning.

Third example: international accounting rules. Evolved among international companies. Companies in developing countries can follow these rules and get better access to markets. But this can be very expensive, keeping poorer countries shut out of international credit. Chinese companies now have, by legend, four different sets of books: for the government, for the firm, for foreign partners, and actual accounts. (Joel's note: people respond to incentives.)

"What are institutions? What differentiates them from organizations, governance, law, regulations and public policy? Do institutions matter? How and why do they matter?"

World Development report 2002. Institutions and Markets. Chap 1

Summary: good institutions (everything from free press to working bankruptcy courts and easy, low-cost business registration) correlate closely with high GDP.

The four main lessons of this chapter are that for effective institution building policymakers need to complement what exists, innovate to suit local conditions, foster open trade and open information exchange, and foster competition among regions, firms, and individuals.

Douglass C. North (1993), Economic Performance Through Time, Nobel Prize lecture, 12/19/93.

It was a mistake to try reading this in a couch instead of an uncomfortable chair. A few quotes that are pleasing in that they suggest a sane understanding of how things work, and yet are uttered by a prize-winning economist:

It is necessary to dismantle the rationality assumption underlying economic theory in order to approach constructively the nature of human learning...

We cannot account for the rise and decline of the Soviet Union and world communism with the tools of neo-classical analysis, but we should with an institutional/cognitive approach to contemporary problems of development ... It is the admixture of formal rules, informal norms, and enforcement characteristics that shapes economic performance. ...

a. Political institutions will be stable only if undergirded by organizations with a stake in their perpetuation.

b. Both institutions and belief systems must change for successful reform since it is the mental models of the actors that will shape choices.

c. Developing norms of behavior that will support and legitimize new rules is a lengthy process and in the absence of such reinforcing mechanisms polities will tend to be unstable.

Lecture handout

"What causes underdevelopment? Which institutions are needed for development?"

Shirley, M. 2004. Institutions and Development: Presidential Lecture at the 2004 Meeting of the International Society for Institutional Economics

Chang, H. J. 2007. Kicking Away the Ladder. Chapter 3

How institutions matter to development: A stark reality

This is a page about how North Korea is dark at night. More interesting to me (because it relates to something I've talked about earlier, California's energy efficiency) is this slide on a related presentation. If you sketch in some best-fit lines, California is clearly emitting much less light at night than it should. Why? Energy-efficient fixtures that don't shed light into the sky? Do they simply need less lit area because Los Angeles is the densest city in the United States?