Apr 20 > So, a few questions regarding the final: > Most importantly, for what am I responsible? It will be cumulative, covering everything from the start of the course. So use the list of topics that I gave you for the second midterm and removed the X's. > Just out of curiousity, have there been any others who expressed interest? > If not, of if there are only a handful, how will the curve work? So far I only have 2 others who've said they'll take it, so it seems that it won't be many. Therefore it can't be curved. I'll just read your answers and give you whatever letter grade seems appropriate. I don't see any other way. Apr 20 > Do you plan to post the past year's final exams on your web page, as you > did for mid-terms? I didn't give final exams in the course before. I've posted all I have. Apr 19 > 1) The topics in the final will be the same as those in the last > mid-term? Or will you provide the questions which you didn't in the > mid-term? I mean , for the last mid-term, you gave us the information > that you didn't require us to do the questions like 'monopoly', > 'Pareto-Criterion', and so on. Will you plan to do the same thing? Good question. No, the final will be cumulative, covering the entire course. The same list of topics, but with nothing crossed out. > 2) Will you have an office hour in the next week as usual? Yes, Tuesday at 9. Apr 19 > How were course grades computed? Rather, what was the scale/ breakdown > for grades? I normalized the scores, computed a weighted average of those using the weights I announced, and then curved those averages setting the median between B+ and A-, more or less. Apr 13 > I was just wondering if you could provide a time frame for when we will > receive our grades, so we can make an informed decision about the final. I hope to have the exam grades by early next week, but I haven't spoken to Bronwen and Sona recently to learn how it's going. As soon as I get the exam grades, I will curve the exam and post them. Then I will calculate course grades from that and post them as well. So I'm hoping that by early next week (late Monday, perhaps?), you will have full information. > If informing us this week is not possible, and I sort of guess that it's > not, do you have any idea about the mean and distribution of the exam? I > am right on the border of an A/A-, and if I got an A on the midterm, I > probably wouldn't have to take the final. A rough estimate based on the > posted answers gives me around a 42 or 43 out of 54. I really don't have any hard information on that. The ones I graded ranged from well below that to somewhat above, but I have no idea whether they were representative. Apr 9 > When we went over the elasticities on taxes and subsidies, I > understood that the quantities stayed the same but that both the tax > and subsidy created price wedges. In the last example, the government > sold bushels of electricity. Why does this change quantity and not > price? (is the government supplying a quantity, thus creating a wedge > in quantities at a lower price?) No, the tax or subsidy does NOT leave quantity unchanged. It changes both the quantities supplied and demanded, but by the same amount. Likewise, when government intervenes in quantity, buying or selling some of the good, this also changes the prices paid by demanders and received by suppliers, but again by the same amount, and changes their quantities by different amounts so as to accomodate what the government buys or sells. Apr 7 > Might you be able to prepare some sample questions for some of the > recent material for which we were unable to have an opportunity to work > through problems (i.e. problem sets). If these topics are to be covered > in a format other than problems (e.g. short answer) that my point may be > moot, but it would be helpful to have an idea of how questions relating > to issues like the leaky bucket will appear. I think the problems that I went through in class on Wednesday pretty well cover it: If the leaky bucket ratio is c and if a project will cause benefits/costs of dYn and dYp for the non-poor and poor respectively, then is the project good or bad from the standpoint of the leaky bucket? That is, if the project benefits the poor and hurts the non-poor, could a better result be achieved with the leaky bucket instead? And if the project benefits the non-poor and hurts the poor, could it be accompanied by use of the leaky bucket in a way that would make both groups gain? I also illustrated that if you calculate the break-even weight on the poor for the leaky-bucket policy, wb=1/(1-c), then you can answer these questions equivalently by simply adding up the dY's weighted by one for the non-poor and wb for the poor. Apr 7 > I am a little unsure what chapters should have been read for this test? Good question. I'd say just chapters 1-4, and 6-7, and in Ch. 7, we really only got to p. 123. Apr 5 > Quick question. I didn't hear you mention elasticities on either the > "not on exam" or "definitely study" list. Do you feel comfortable > designating it on one side of the fence or the other? Yes, I should have mentioned it as "definitely study." Apr 2 > Who owns the property where the new factory is to be located, and what > is its value? As usual, you can make whatever assumption you think is reasonable. Myself, I'm assuming that the property has no value at all, since it is currently sitting there as an eyesore (I presume -- I've never seen a defunct boring factory) and nobody seems to want it. Apr 2 > We have the following question about the case. When you write that > "Softwick demands a guarantee that they will not be sued," are they > demanding that the city government not sue them, or are they somehow > demanding that the city government prevent Softwick citizens from suing > them (in which case, would the city government have to compensate the > worker's family for the value of the lost life instead of the company?). > > Hope this makes sense, Yes, it makes sense. But I don't know the answer. I just wanted to ignore the possibility of lawsuits (even though I don't know if that is really possible, legally). If the city has to compensate anybody for lost life, that won't matter for the bottom line, right? Since such transfers within the city will cancel out. Apr 2 > We have a couple of questions regarding the case exercise: > > 1) With regard to the market for building materials, are the 300 tons > that are required for construction bought all at once or spread out > throughout the three years of construction? I suppose either is possible, but my own assumption was that they were spread out over the three years. > 2) The general data says that annual income a year ago was $120 > billion. However, the introduction says that everyone works at a wage > of $20 or $22.50 if they are in construction. Should we assume that > this added income comes from outside of the city? or are some of the > workers earning more than the $20 wage (or $22.50)? The wage will only give you labor income. A community's income, like GDP, comes from other sources than that -- most obviously capital ownership. I'm not sure how this matters for any of your calculations, but that was my thinking in stating that figure. Apr 1 > Does the factory continue to smell bad after the 30 years of operation? I hadn't thought about that. Logically, it should probably stop smelling then, but then the price of land nearby should depend on how many years it has left to smell, which I didn't allow for in my regression. So it might be simplest to assume that it keeps on smelling bad forever, even though that is implausible. I don't think it will matter much either way, except for the complexity of the calculations. You can do it either way. > Are property taxes paid by the factory before it is done being built? I don't know what is actually done. I had supposed that the taxes would start only after it is built. Apr 1 > Are we to assume that the 100 construction workers hired to complete > the factory project are drawn from the population of unemployed > workers? Or > are they currently employed by construction firms and therefore not > part of the unemployed population? I would assume that they come from the same local labor market as everyone else, so that either they were unemployed themselves, or their hiring would open up jobs for someone else. So that, in effect, they do come from the ranks of the unemployed. If you think that is implausible, you are welcome to assume something else. But they definitely should come from some local labor market, not from outside. As for the premium they must be paid to bear the risk of fatal accident, I think of that as just a premium that must be paid to anybody, over the standard $20 an hour, for this level of risk. Mar 31 > Question on the computations > For the construction workers, the 6% translates into this example. I > have > a hundred workers, I draw one name out of the hat, and then that person > has a 6% chance of dying or ie, I have to pick one of the six out of one > hundred cards, for that person to die. Is that what the 6% is saying? Yes, I guess so. It is NOT saying that every worker individually has 6% chance of dying, but that the group of them collectively has a 6% chance that just one of them will die. > If so then it is the equivilent fraction of 1 in X which is equal to > this example. so X times 2.50 is the value of a life correct? Yes, except that the 6% probabiliy is for a year, and the 2.50 is for an hour. Mar 30 > My group doesn't know how to work with logs. Could you give us a > primer? > Specifically, how do you convert from a natural log to a non-log number? First, what you DON'T need to know: The logarithm of a number, x, to some base "b", is the power of b that equals x. In other words, if y=logb(x), then b raised to the power y equals x. The two most common forms of logarithm are to the base 10 and to the base "e", where "e" is a particular number (I don't recall what it is -- I've never needed to know it) that has nice properties (I don't recall these either). Logarithms to the base e are called "natural logarithms," and sometimes denoted "ln" instead of "log". Logarithms are only defined for arguments that are positive numbers, since you cannot get a negative number by raising anything to a power. On the other hand, the logarithm of a positive number will itself be positive, negative, or zero depending on whether the argument is larger than, smaller than, or equal to one (since anything to the power zero is one). The inverse of a logarithm is the "exponential", which merely means the number that you get when you raise that special number e to some power. That is, exp(x) means e raised to the power x. From these two definitions, it follows that if y=ln(x), then x=exp(y). The exponential is defined for arguments both positive and negative, but it only takes on values that are positive (less than one if the argument is negative). One reason that logarithms are useful is that they represent simply relationships that are curved instead of straight. Sometimes from the logic of a problem, that is more realistic. So quite often when econometricians try to fit an equation to some data, they find that logs of the variables fit better than just the variables by themselves. Now, what you DO need to know: Presumably you will be using a spread sheet and/or a calculator. These routinely have both a "ln" function and an "exp" function. You can just use these to convert back and forth. To check that you are using them right, I suggest that you test by picking a number, calculate its ln(), then calculate the exp() of that and see if you get back the original number. I'm not at all sure you will need this next bit, but in case you do: For any positive x and y: ln(xy)=ln(x)+ln(y). For any x and y (regardless of sign): exp(x+y)=exp(x)exp(y) Mar 30 > Are the building materials for construction coming from outside the > town? No, I meant them to be purchased on the local market, adding to the demand there. > You write the equation with "log" but then say "(natural) logarithm. > Which is it, log or ln? I meant it to be the natural logarithm and didn't realize that had to be written "ln." Change it to "ln" if that will clarify. Mar 30 > 1) Should we assume wages get adjusted each year for inflation? Yes, definitely. That is by far the easiest, and I see no reason to assume otherwise. > 2) What was the former property value of Humdrum Boring Works? You > state that the current property value is $800 million EXCLUDING the > old company. If we want to present an option to turn Softwick away, > we'd need to calculate the loss in city revenue of not replacing a > company. We'd also like to compare whether Softwick has a greater > property value than the old company. Given that the property has remained empty for a year and that this is the only use of it that has appeared so far, I would not assume that it has much value. I'd probably take it to be zero, myself, but you may prefer to assume something positive. > 3) What is the "normal turnover rate?" We'd like to calculate what > percentage of the 2000 member labor force leaves each year for > retirement, death, winning the lottery, maybe moving, etc. so we can > calculate how many of the 900 unemployed would get absorbed into the > local labor market through frictional unemployment. Turnover means, to me, that people both enter and leave the labor force. You can assume whatever rate seems plausible to you, but I don't see how it will matter for the overall results. > 4) Are we assuming there is no environmental regulation (EPA, or > state) in Humdrum, WC that would prohibit discharge of the stated > concentration/quantity of hexavalent yuck? That's what I had in mind. No such regulation is mentioned. Feb 21 > We were confused about the following aspect of the problem set. We were > tackling the problem in four different ways and then we were going to > campare each of the outcomes. > > 1. Renovate the theater but don't build parking. > 2. Renovate the theater and build free parking. > 3. Renovate the theater and charge for parking. > 4. Do nothing and shop at the mall. Those sound like good options to look at, although I admit I haven't looked at them all myself. > The problem is that when we go through 1, we make the assumption that > you can fill the theater to capacity without building parking (i.e. people > walk). Then later, that means that there is really no benefit to > parking > at all because we've already stated that we can fill the theater to > capacity without parking. That doesn't seem quite right, since the people who pay for parking would surely have been willing to pay more, some of them, so that there is consumer surplus for them from having the parking structure, even if it doesn't alter the viability of the theater. But from the writeup, I presumed also that the inconvenience of parking would be enough to keep the theater from filling up. Of course, I don't know how I would quantify that, since the case doesn't include information on that. > Now we're wondering if we were wrong to try and look at all four > options. Is it okay if we just pick one of them? Well, implicitly, number 4 is the base against which the others are compared, so you do it whether you want to or not. And yes, it's fine if you pick one, though I might feel that you had undermined some of the point of the case if you completely ignored the parking structure. Feb 20 > 1) We are able to identify all the costs and benefits enumerated in the > text of the case exercise, however, this then seems to make the exercise > one of > addition and subtraction -- we feel we are missing the economics of the > exercise - related to supply of theatre productions (total cost and > marginal cost), consumer surplus, and the object of any sensitivity > analyses. Well, a lot of it IS addition and subtraction, but the rest of the questions that you ask make it clear that you are looking at the economics of this, so I wouldn't worry about that. Your other questions mostly have to do with the fact that the case doesn't spell out everything so that you can be absolutely sure what is going on. That is intentional. You need to make assumptions, such as that some of what you don't know is similar to what you do know, that functional forms are either straight lines or constant elasticity depending on what is either more tractable or more plausible, that marginal costs of some things are constant and perhaps even zero, etc. Note too that working with the elasticities of curves is mostly useful when you don't know much about the curve, as when you only know one point. If you think you know two points on the curve, then that already defines the elasticity, at least between those points, and it also makes it rather unnecessary. If you assume the curve is linear, for example, then two points are enough to tell you the whole thing. > We are also unclear how to graphically > represent ALL the benefits and costs (i.e.the market of theatre > productions do not capture ALL the benefits and costs) I hadn't supposed that you would use graphs for everything. I'm not even sure that is possible. If a graph can help to illustrate some of what you do, then by all means use it, but graphs are not necessary, and sometimes they are not even helpful. Feb 20 > ....Is this a waste of time? Are we barking > up the wrong tree? We are very uncertain as to the level of detail we > need to go into in this regard. Since I want to share any answers that I give with others, I think I will have to decline to answer this. I couldn't have my answer make much sense without showing them your question, which would in turn tell them more than I probably should. If you are on the right track, I don't want to help them inappropriately. If you are on the wrong track, I don't want to mislead them. I'll just say this: I don't expect everybody to necessarily handle this case in the same way. Feb 20 > We are a little confused about the different property valuations > reported by different parties. It is likely to be true that the different witnesses may provide conflicting information. You need to use your judgement in deciding what they meant and whether to trust their information. > Beancounter describes the city's assessed valuation > as being 1/2 of the market value. Is this consistent with Whatnow's > reported valuations or do those represent market value? My understanding is that an appraiser's job is to try to determine the market value, so yes, I would take Whatnow's evidence as being about market value. > And how about for > the commercial properties- Does the average value of $60,000 indicate > market value? That's what it sounds like to me. How else would you interpret it? Feb 18 > OK, we think we understand it now, but I'm going ot tell your > interpretation to make sure it's right: every year, including the first > year, we should include as a cost 3% of the one-time costs, not the full > amount. Yes. And then, because all years are the same, you never need to look at more than one. Feb 18 > We were wondering if the results of the consulting job done by Just Ask > Us indicate that the theater could fill 1200 seats per week at $1 or 1200 > seats per event (in other words, 12000 theater goers per week) at the > price of $1. And of course, likewise, would they fill 500 seats per week > at $5 or would that be 500 seats per event (5000 per week) at the price > of $5. I read it as saying that they'd get 1200 per event (and thus 12,000 per week), etc. Feb 18 > Also, we are confused re: the cost of supplying water to the theatre - > should we make assumptions about whther or not it costs the city money > to provide the water for free? On that I think you should use your judgement, and defend it. DOES it cost a city to provide water for free? Feb 18 > my group is experiencing some confusion re: perpetual flows and one-time > costs, namely that we don't fully understand the former and how it > affects > the latter. How do assess the costs of the project without knowing how > mnay years to allow for time discounting? Please explain this portion > in a little more detail. Yes, I realize that is confusing, and I don't want it to be. Basically, I do NOT want you to have to worry about discounting on this assignment at all. So you should put all costs and benefits into the form of costs and benefits per year, and work with that. For any cost or benefit that occurs every year, you should assume (however unrealistically) that they continue every year foreever, so that you don't have to worry about that at all. And for all one-time costs and benefits, just multiply by the interest rate, 3%, to make them too ongoing perpetual flows. In the case of a one-time cost, this is equivalent to borrowing to cover it and then paying only interest on that debt forever, sort of the way that I seem to handle my credit card debt. In the case of a one-time benefit, this is equivalent to "annuitizing" it for a person with an infinite life span. But you shouldn't have to worry about either of these interpretations. Just do it as I've described, and don't bother mentioning issues of time in your writeup. This aspect of the problem should therefore be very simple, and if it's not, then you are misunderstanding what I intend, and you should ask again.