|
|
|
|
| I have finished posting answers and hints to the review questions on the three sections (introduction, data of macro, the classical model) that will be covered on the mid-term. | ||||||
| The class for October 7 will now NOT be cancelled. October 9 is still cancelled, as I will be giving a talk at the Board of Governors of the Federal Reserve. | ||||||
| The due date for problem set 3 has been pushed back one week to Friday October 11. | ||||||
I am trying to maximize my availability before the mid-term.
This is not easy to do:
|
|
|
| If you visit the page on review questions for the classical model, you will now find links to answers for almost all of the review questions. I hope you find this helpful in revising for Friday's exam. I am working on links for the review questions in the section on the data of macroeconomics, and will get them up as soon as I can. |
|
|
| I have posted to the website 20 or so new review questions for the classical model. You will notice that the website has been restructured. Select the option "Review Questions" from the contents menu on the left. This will lead you to review commentaries and questions for the three sections covered in Firday's mid-term exam. These review sections can also be accessed directly from the introductory page to each section. |
| On Wednesday we will have a review session in class. We will go through a sample of questions from the reviews. As there are 40+ questions in the reviews, we will only cover a fraction of them. I will leave it to your requests as to which questions we look at. |
| A number of students have not attended class since the first problem set was handed in, and I still have them. While I have no explicit attendance policy, please be aware that my consulting fees for private repetition to absentee students of material covered in lectures are too exorbitant for you to afford. |
|
(Section B) |
| Please find attached my answers to problem set 1. |
|
(Section B) |
| I look forward to meeting you tomorrow even if I, like you, would rather be sleeping. It gets worse. |
| We will meet tomorrow at 8:30am. I will begin the course at the beginning, but we need to play catchup with the other class for two reasons. First, the later class reamins overloaded, and we need to switch some students from there to the early session. Second, having both classes progress at the same rate helps me out enormously: I have only a single grader, I can send common emails on subject matter to both sections; I only have to maintain one web site with announcements, new reading materials, release of answers to problem sets, and so on. More important, when you come to see me, I know what I've taught you and what I haven't taught you. |
| So, to play catchup, on Monday and Wednesday next week (and possibly Friday), we will start early(!) meeting from 8:00 am to 9:20pm. The original section has a problem set due on Monday 9th Sept. Your will need to hand in this problem set on Monday 16th September. By then, we will be back to normal. |
|
(Section A) |
| The new class section has now officially been added. It is to be held from 8:30am to 9:20am, MWF in Room 226B. If you have been attending the 9:30 class up to now, but would either like to switch to 8:30 or need to switch to 8:30 because you were waitlisted, please continue to attend the 9:30 session for a few more days. For the first few classes I will be playing catchup. I will go a bit faster and meet from 8:00am to 9:20am on Monday and Wednesday next week (possibly also on Friday). By the end of next week, I anticipate that both classes will be at the same point. After that, everything will go smoothly. | ||||||||
| Revision Summary | ||||||||
| Here is a summary of where we are so far. Most of the summary is in the form of questions. If, after revising the material, you can answer all these questions without much though, AND you can do the problems in set 1, you should feel confident that you have mastered the material. | ||||||||
| I recommend you take time now to prepare asnwers to these questions -- they would make useful summary notes in your own words for later revision. In a few weeks, pull the questions out again and see if you can still answer them all easily without looking at your notes. Following these suggestions will also make your life more exciting and fulfilling. | ||||||||
| 1. Introductory Talk | ||||||||
| There were a bunch of readings, about what makes a good model; the size of shopping carts, and eskimo words for snow, that were really designed to get your thinking juices flowing. I hope you enjoyed them. | ||||||||
| The serious points we made in this section, however, concerned the difficulties that economists have in developing, comparing, and testing theories. For obvious reasons, economists aren't often allowed to design experiments with peoples lives, and this creates problems that researchers in areas such as medicine and physics don't have. | ||||||||
| You should now be familiar with the following concepts: | ||||||||
| a) The identification
problem.
b) The sample selection problem. c) The search for natural experiments. |
||||||||
| For each of these, you should be able to describe in general terms what they are, and why they matter. You should also be able to provide a couple of examples of each. One of my favorite exam questions is to ask students to prove they understand a concept by providing an example that was NOT presented by me or discussed in class. You might like to tackle this question now. | ||||||||
| Over the next couple of months you should also begin to relate debates about theories to these concepts, and build up a number of macroeconomic examples of each concept. Overall, if you were asked why it is that whenever you put two economists in a room you always get three opinions, you should be able to give a coherent explanation. | ||||||||
| 2. The Data of Macroeconomics | ||||||||
| We looked at 3 pieces of information: gross national product, inflation and unemployment. For each of these, we asked the following questions: | ||||||||
| a) What would we like it to
measure?
b) How is it actually measured? c) How has the variable behaved over time? |
||||||||
| Overall, we find that the official statistics don't measure exactly what we care about (or exactly what our economic theories may be about), and even what they set out to measure is done with error. Despite these inadequacies, however, the data remain extremely useful. | ||||||||
| You should be able to answer the following questions: | ||||||||
a) GDP
|
||||||||
b) Inflation
|
||||||||
c) Unemployment
|
|
|
| Some of you have already heard that we have been unable to find a new classroom to clear the waitlist. We are therefore in the process of setting up a new section of the course, which will be taught from 8:30am to 9:20 to accommodate people on the waitlist. I anticipate that this will also be in room 235A. I hope that the new section will be established officially tomorrow. |
| Connie Angermeier, the department coordinator in SDS, is in charge of setting this up. She will also handle transfers of people from the waitlist to the new section, as well as registration of people who were unable to join the waitlist. She can be reached at <cla2@andrew.cmu.edu>. I hope this can all be done smoothly over the next couple of days. |
| If you know of students who planned to register but who are not on the waitlist, perhaps you would be kind enough to let them know about the new section and invite them to contact Connie. |
| I appreciate that a new time slot, especially one so early, is not the ideal solution (some of you may have a conflicting class at that time, in which case you should talk to Connie). Unfortunately, this is the best I have been able to do. |
| I must also ask you all to respect the following rule: Students registered for the 8:30am class must attend the 8:30 class, not the 9:30am class, unless prior arrangement has been made with me (through Connie). We can only consider exceptions for students already enrolled in a class at that time, and only insofar as room capacity allows us to do so. |
| If there are any students registered for the 9:30 class who would prefer to attend at 8:30, you are welcome to do so. Please let me know, however, so I can be aware of changing capacity at the 9:30 session. If you decide to make the switch, please commit to that for the remainder of the semester. |
| Finally, I have lost a lot of time with this today, so the summary of the key points on the Data of Macroeconomics has not been prepared. I will do that in the morning and email it out to you. |
| Thanks again for your patience. |
|
|
| 1. Just in case some people weren't in class today, we will be meeting on Friday in room 235A. There are no more chairs, but the layout is a little better. I remain hopeful we can do better next week. |
| 2. In case you're not on the ball, I wanted to remind you that the first problem set is due at the beginning of class on Monday 9 September. You should have enough material covered to work on it over Labor Day weekend. Lucky for you. |
| 3. Students who are attending but either on the waiitng list or waiting to transfer form intermediate macro are of course not expected to do the assignment. If we can get you in the course next week, we'll adjust the weights on the remaining problem sets. |
|
|
| 1. Classroom |
| As some of you may have noticed, the classroom is a little small for our needs. I have put in a request for a larger room, where I hope everyone will be able to see the board and those currently waitlisted can register for the course. Apparently, our administration does not do room changes in the first week, period. I will try the strategy of claiming that having people sit on the floor in front of the door is a fire risk -- that usually works. My colleague Steven Klepper has also kindly offered to walk the halls looking for an empty room at 9:30 Wednesday, and if he finds one we will occupy it by force. As soon as I have any news I will let you all know. In the meantime, I am afraid, we are just going to have to put up with this. |
| 2. The value of schooling |
| The paper that conducted the
study of the effect of compulsory schooling was written in 1991 by Josh
Angrist and Alan Krueger and can be found online here.
One interesting thing about this paper. Two years later a paper criticized
the natural experiment [John Bound, David A. Jaeger, Regina Baker in "The
Cure Can Be Worse than the Disease: A Cautionary Tale Regarding Instrumental
Variables." NBER
working paper t0137, June 1993].* We saw in class that the idea
of the natural experiment in this case was to find some way to split the
sample of students up into those who had more and those who had less, schooling,
in a way that was uncorrelated with their ability. Angrist and Krueger
came up with birth date, which can reasonably be argued to be unrelated
to ability. They then argued that quarter of birth should affect wages
only through its effect on the amount of education people had. Thus, looking
at the effect of birth quarter on wages should help measure the effect
of an extra year of education on wages that does not result from ability
differences (the actual technique used is known as intrumental variable
regression, and requires advanced statistical knowledge).
The interesting claim in Bound et al. is that the quarter of birth may also have a direct effect on wages, independently of its effect on the amount of schooling kids get, because it is correlated with ability. If this is the case (and they provide statistical -- again technical -- evidence that it is), then using quarter of birth no longer gives you the effect of extra education not caused by differences in ability: it gives you the effect of extra education caused by quarter of birth PLUS some direct effects of quarter of birth on wages. That was not what Angrist and Krueger were after at all! Here are some reasons cited by Bound et al. for a direct effect. a) Quarter of birth is related to age at entry, and many educators believe that age of entry affects performance. b) There is some evidence that individuals born early in the year are more likely to be mentally retarded, as well as mixed evidence on variations in IQ by birth date. c) There are regional variations in birth seasons. Those in the south are less likely to give birth in February (I presume because it is so hot in May). Birth season will then be correlated with regional variations in ability, which also exist. Bound et al. conclude that we cannot trust the Angrist-Krueger results. This is a real shame -- Angrist and Krueger came up with what appeared to be a very nice "natural experiment". If an idea as neat as their's doesn't pass muster as a natural experiment, what will? |
| * A note here -- these are technically difficult papers. The links are really put here for completeness. The links will also only work smoothly from a CMU domain IP address. |
| 3. And finally . . . |
| Who the hell is Julian Huxley? Your answer is here. |