.
1. Driving early this morning, I listened to NPR's Morning News. The news said: "The library of the University of Illinois at Urbana-Champaign is celebrating their acquisition of it's 10 millionth volume." The campus' News Bureau said: "[This library] is the largest public library in the world." Wow. Then I realized how so many students really don't appreciate this superfacility. I know one student who got his Ph.D without ever visiting library! I know some advance graduate students who have no idea how to find books in the main stacks. I know many students who never check books online. Who don't know that there is this thing called "interlibrary loan". Who don't know that the library has fun stuff: novels, videos, etc. Who don't know there are interesting seminars every now and then in the underground level of the main building. Who don't know there are art movie series and talks once and a while. Very sad. For sure this -- our beloved library-- is the biggest thing I will miss when I go home. Well, some people just don't care. Some people don't even read.

2. Good, the former governor of Illinois, George H. Ryan failed to win the 2003 Nobel Peace Prize. How come someone so corruptive like him ever got promoted in the first place?

3. While in West, the craziest state of America, California has elected it's new governor. He's the Terminator: Arnold Schwarzenegger. What to say?

4. Had a meeting with JBB. I proposed my idea on capturing the warm glow effect. Basically what I want to do is:

4.1) Settle on the definition of "warm glow" that I will analyze. My standpoint is that agents prefer regarding themselves as socially responsible individuals. So this is not exactly similar to "altruism" where an agent's utility function include that of another's. In fact, I believe, an agent can get a moral satisfaction simply from the act of giving and not necessarily out of, say, environmental concern -- in a case of environmental public good.

4.2) It follows, with regards with our survey, I am offering three propositions:
4.2.1. If an agent thinks that the harbor pollution is not important at all, while she reveals a higher-than-average willingness-to-pay for cleanup, then she has purchased some amount of moral satisfaction, i.e. warm glow. 4.2.2. If an agent agree that the harbor area is environmentally safe, while she reveals a higher-than-average WTP for cleanup, then she has purchased some amount of moral satisfaction.
4.2.3. The relevant value to be associated with an environmental improvement is therefore the total value minus warm glow value.

4.3) To operationalize those propositions, I'm offering the following steps:
4.3.0. Derive the model: Direct Utility Function to Indirect UF to Compensating Surplus. 4.3.1. Identify who reveals warm glow effect using the previously calculated WTPs as filter. 4.3.2. Create dummy/effect to distinguish between households w/ and w/o warm glow. 4.3.3. Interact the dummy/effect with choice-specific variables to avoid cancellation. 4.3.4. Reestimate the WTPs with full model that includes the added, warm-glow dummy/effect interaction. 4.3.5. Calculate the difference between WTPs here and the one calculated before. Hypothesis: The new WTPs are higher than the old ones. 4.3.6. Test the significance of the difference, if any. This difference is the warn glow effect. 4.3.7. Assess the warm glow effect based on geographic, income, or ethnic groups.

JBB raised an important concern: Is it alright to use WTPs as the filter for categorizing the respondents, where the WTPs were calculated using previous estimation? That is, he was concerned with the problem of endogeneity. My first reaction was "maybe it's OK". Why? Because we won't use the same WTPs in the next estimation: they are just some filtering mechanism. We will reestimate everything and recalculate WTPs with warm glow dummies taken into account. Need to dig further on this issue.

Alternatively, JBB suggested, try to create new indicators to serve as the proxy of warm glow. Still vague, but he was wondering if we can somehow lump some attitudinal statements and generate a composite index. We'll see how it plays out.