What’s wrong with my program?My issue: My program does not work for the input of 1,000,000 and I am not sure what I’m doing wrong. When I try to input 1,000,000 it stops at around 180,000Write a C program that will create threads (Prime Thread and Print Thread) that will find and print the prime numbers between 0 and an integer number that the user provides. There is no limit on the number of threads and you can create as many threads as you need.The prime numbers will be stored in a shared buffer named “Prime”. The threads will take turns between finding the prime numbers and printing the found numbers.MAINThe main will get a number from the user and store it in a variable named “Max_Num” and will create the threads (Prime Thread and Print Thread) to find and print the prime numbers.Prime Thread and Print ThreadThe Prime threads will find the prime numbers and will store them in a global shared variable “Prime”. The Prime threads will continue to produce prime numbers until the buffer is full then will wait until the Print threads empty the buffer and print it to the screen.You can use any synchronization technique (lock, mutex, semaphore) we discussed in class. You may need to use a Mutex lock “Lock_Prime” to ensure that only one thread accesses the “Prime” buffer at a time. Your program should run reasonably fast for any given number.-Your program should be able to adjust the number of threads according to the input size.-You can use any synchronization technique. Code must be well documented with plenty of comments.-Your code should work correctly with any input.-While some delay is accepted, results should be ready within a reasonable time.-Your program should be able to accept inputs of 10000, 100000, and 1000000. For 1000000 a reasonable execution time is < 5 minutes.- The waiting for the completion of threads should occur at the end i.e. threads need to be created before any of the joining happens.- The program should print the amount of prime numbers within the given range, as a checkExample: N = 10Acceptable: 2,3,5,7 3,2,5,7 5,7,2,3Not Acceptable: 2,2,3,5,7 2,3,5
1DATA SET HANDBOOKIntroductory Econometrics: A Modern Approach, 7eJeffrey M. WooldridgeThis document contains a listing of all data sets that are provided with the seventh edition ofIntroductory Econometrics: A Modern Approach. For each data set, I list its source (whereverpossible), where it is used or mentioned in the text (if it is), and, in some cases, notes on how aninstructor might use the data set to generate new homework exercises, exam problems, or termprojects. In some cases, I suggest ways to improve the data sets.Special thanks to R.G. Wooldridge for updating the page numbers for the 7th edition.401KSource: L.E. Papke (1995), “Participation in and Contributions to 401(k) Pension Plans:Evidence from Plan Data,” Journal of Human Resources 30, 311-325.Professor Papke, of Michigan State University, kindly provided these data. She gathered themfrom the Internal Revenue Service’s Form 5500 tapes.Used in Text: pages 62, 76, 133, 169, 212-213, 656-657Notes: This data set is used in a variety of ways in the text. One additional possibility is toinvestigate whether the coefficients from the regression of prate on mrate, log(totemp) differ bywhether the plan is a sole plan. The Chow test (see Section 7.4), and the less restrictive versionthat allows different intercepts, can be used.401KSUBSSource: A. Abadie (2003), “Semiparametric Instrumental Variable Estimation of TreatmentResponse Models,” Journal of Econometrics 113, 231-263.Professor Abadie, now at MIT, kindly provided these data. He obtained them from the 1991Survey of Income and Program Participation (SIPP).Used in Text: pages 160-161, 178, 217-218, 258-259, 276, 292, 528Notes: This data set can also be used to illustrate the binary response models, probit and logit, inChapter 17, where, say, pira (an indicator for having an individual retirement account) is thedependent variable, and e401k [the 401(k) eligibility indicator] is the key explanatory variable.2ADMNREVSource: Data from the National Highway Traffic Safety Administration: “A Digest of StateAlcohol-Highway Safety Related Legislation,” U.S. Department of Transportation, NHTSA. Iused the third (1985), eighth (1990), and 13th (1995) editions.Used in Text: not usedNotes: This is not so much a data set as a summary of so-called “administrative per se” laws atthe state level, for three different years. It could be supplemented with drunk-driving fatalitiesfor a nice econometric analysis. In addition, the data for 2000 or later years can be added,forming the basis for a term project. Many other explanatory variables could be included.Unemployment rates, state-level tax rates on alcohol, and membership in MADD are just a fewpossibilities.AFFAIRSSource: R.C. Fair (1978), “A Theory of Extramarital Affairs,” Journal of Political Economy 86,45-61, 1978.I collected the data from Professor Fair’s web cite at the Yale University Department ofEconomics. He originally obtained the data from a survey by Psychology Today.Used in Text: not usedNotes: This is an interesting data set for problem sets starting in Chapter 7. Even thoughnaffairs (number of extramarital affairs a woman reports) is a count variable, a linear model canbe used to model its conditional mean as an approximation. Or, you could ask the students toestimate a linear probability model for the binary indicator affair, equal to one of the womanreports having any extramarital affairs. One possibility is to test whether putting the singlemarriage rating variable, ratemarr, is enough, against the alternative that a full set of dummyvariables is needed; see pages 239-240 for a similar example. This is also a good data set toillustrate Poisson regression (using naffairs) in Section 17.3 or probit and logit (using affair) inSection 17.1.AIRFARESource: Jiyoung Kwon, a former doctoral student in economics at MSU, kindly provided thesedata, which she obtained from the Domestic Airline Fares Consumer Report by the U.S.Department of Transportation.Used in Text: pages 476, 488, 557, 557Notes: This data set nicely illustrates the different estimates obtained when applying pooledOLS, random effects, and fixed effects.3ALCOHOLSource: Terza, J.V. (2002), “Alcohol Abuse and Employment: A Second Look,” Journal ofApplied Econometrics 17, 393-404.I obtained these data from the Journal of Applied Econometrics data archive athttp://qed.econ.queensu.ca/jae/.Used in Text: page 600APPLESource: These data were used in the doctoral dissertation of Jeffrey Blend, Department ofAgricultural Economics, Michigan State University, 1998. The thesis was supervised byProfessor Eileen van Ravensway. Drs. Blend and van Ravensway kindly provided the data,which were obtained from a telephone survey conducted by the Institute for Public Policy andSocial Research at MSU.Used in Text: pages 195, 217, 259-260, 598Notes: This data set is close to a true experimental data set because the price pairs facing afamily were randomly determined. In other words, the family head was presented with prices forthe eco-labeled and regular apples, and then asked how much of each kind of apple the familywould buy at the given prices. As predicted by basic economics, the own price effect is negative(and strong) and the cross price effect is positive (and strong). While the main dependentvariable, ecolbs, piles up at zero, estimating a linear model is still worthwhile. Interestingly,because the survey design induces a strong positive correlation between the prices of eco-labeledand regular apples, there is an omitted variable problem if either of the price variables is droppedfrom the demand equation. A good exam question is to show a simple regression of ecolbs onecoprc and then a multiple regression on both prices, and ask students to decide whether theprice variables must be positively or negatively correlated.APPROVALSource: Harbridge, L., J. Krosnick, and J.M. Wooldridge (2016), “Presidential Approval andGas Prices: Sociotropic or Pocketbook Influence?” In New Explorations in Political Psychology,ed. J. Krosnick. New York: Psychology Press (Taylor and Francis Group)Professor Harbridge kindly provided the data, of which I have used a subset.Used in Text: 365, 393, 424ATHLET1Sources: Peterson’s Guide to Four Year Colleges, 1994 and 1995 (24th and 25th editions).Princeton University Press. Princeton, NJ.The Official 1995 College Basketball Records Book, 1994, NCAA.41995 Information Please Sports Almanac (6th edition). Houghton Mifflin. New York, NY.Used in Text: page 661Notes: These data were collected by Patrick Tulloch, an MSU economics major, for a termproject. The “athletic success” variables are for the year prior to the enrollment and academicdata. Updating these data to get a longer stretch of years, and including appearances in the“Sweet 16” NCAA basketball tournaments, would make for a more convincing analysis. Withthe growing popularity of women’s sports, especially basketball, an analysis that includessuccess in women’s athletics would be interesting.ATHLET2Sources: Peterson’s Guide to Four Year Colleges, 1995 (25th edition). Princeton UniversityPress.1995 Information Please Sports Almanac (6th edition). Houghton Mifflin. New York, NYUsed in Text: page 661Notes: These data were collected by Paul Anderson, an MSU economics major, for a termproject. The score from football outcomes for natural rivals (Michigan-Michigan State,California-Stanford, Florida-Florida State, to name a few) is matched with application andacademic data. The application and tuition data are for Fall 1994. Football records and scoresare from 1993 football season. Extended these data to obtain a long stretch of panel data andother “natural” rivals could be very interesting.ATTENDSource: These data were collected by Professors Ronald Fisher and Carl Liedholm during a termin which they both taught principles of microeconomics at Michigan State University. ProfessorsFisher and Liedholm kindly gave me permission to use a random subset of their data, and theirresearch assistant at the time, Jeffrey Guilfoyle, who completed his Ph.D. in economics at MSU,provided helpful hints.Used in Text: pages 110, 146, 193-194, 216Notes: The attendance figures were obtained by requiring students to slide their ID cardsthrough a magnetic card reader, under the supervision of a teaching assistant. You might havethe students use final, rather than the standardized variable, so that they can see the statisticalsignificance of each variable remains exactly the same. The standardized variable is used onlyso that the coefficients measure effects in terms of standard deviations from the average score.AUDITSource: These data come from a 1988 Urban Institute audit study in the Washington, D.C. area.I obtained them from the article “The Urban Institute Audit Studies: Their Methods andFindings,” by James J. Heckman and Peter Siegelman. In Fix, M. and Struyk, R., eds., Clear and5Convincing Evidence: Measurement of Discrimination in America. Washington, D.C.: UrbanInstitute Press, 1993, 187-258.Used in Text: pages 732, 738, 741BARIUMSource: C.M. Krupp and P.S. Pollard (1999), “Market Responses to Antidumpting Laws: SomeEvidence from the U.S. Chemical Industry,” Canadian Journal of Economics 29, 199-227.Dr. Krupp kindly provided the data. They are monthly data covering February 1978 throughDecember 1988.Used in Text: pages 349-350, 359, 363, 407, 410-411, 422, 424, 631-632, 639Note: Rather than just having intercept shifts for the different regimes, one could conduct a fullChow test across the different regimes.BEAUTYSource: Hamermesh, D.S. and J.E. Biddle (1994), “Beauty and the Labor Market,” AmericanEconomic Review 84, 1174-1194.Professor Hamermesh kindly provided me with the data. For manageability, I have includedonly a subset of the variables, which results in somewhat larger sample sizes than reported forthe regressions in the Hamermesh and Biddle paper.Used in Text: pages 231, 259BENEFITSSource: I collected these data from the old Michigan Department of Education web site, whichused to have an annual Michigan Schools Report. I used data on most elementary schools in thestate of Michigan for 1993. I dropped some schools that had suspicious-looking data.Used in Text: 218Notes: This is an elementary school-level version of MEAP93, which contains data for highschools.BIG9SALARYSource: O. Baser and E. Pema (2003), “The Return of Publications for Economics Faculty,”Economics Bulletin 1, 1-13.Professors Baser and Pema kindly provided the data.Used in Text: not used6Notes: This is an unbalanced panel data set in the sense that as many as three years of data areavailable for each faculty member but where some have fewer than three years. It is not clear thatsomething like a fixed effects or first differencing analysis makes sense: in effect, approachesthat remove the heterogeneity control for too much by controlling for unobserved heterogeneity –
which, in this case, includes faculty intelligence, talent, and motivation. Presumably, thesefactors enter into the publication index. It is hard to think we want to hold the main factorsdriving productivity fixed when trying to measure the effect of productivity on salary. PooledOLS regression with “cluster robust” standard errors seems more natural.On the other hand, if we want to measure the return to having a degree from a top 20 Ph.D.program then we would want to control for factors that cause selection into a top 20 program.Unfortunately, this variable does not change over time, and so FD and FE are not applicable.One could include the top 20 dummy variable in a correlated random effects analysis.BWGHTSource: J. Mullahy (1997), “Instrumental-Variable Estimation of Count Data Models:Applications to Models of Cigarette Smoking Behavior,” Review of Economics and Statistics 79,596-593.Professor Mullahy kindly provided the data. He obtained them from the 1988 National HealthInterview Survey.Used in Text: pages 16, 58, 109, 145, 172, 178, 181-182, 251-252, 504BWGHT2Source: Dr. Zhehui Luo, a professor of epidemiology and biostatistics at MSU, kindly providedthese data. She obtained them from state files linking birth and infant death certificates, andfrom the National Center for Health Statistics natality and mortality data.Used in Text: pages 178, 217Notes: There are many possibilities with this data set. In addition to number of prenatal visits,smoking and alcohol consumption (during pregnancy) are included as explanatory variables.These can be added to equations of the kind found in Exercise C6.10. In addition, the one- andfive-minute APGAR scores are included. These are measures of the well being of infants justafter birth. An interesting feature of the score is that it is bounded between zero and 10, makinga linear model less than ideal. Still, a linear model would be informative, and you might askstudents about predicted values less than zero or greater than 10.CAMPUSSource: These data were collected by Daniel Martin, a former MSU undergraduate, for a finalproject. They come from the FBI Uniform Crime Reports and are for the year 1992.Used in Text: pages 128-1297Notes: Colleges and universities are now required to provide much better, more detailed crimedata. A very rich data set can now be obtained, even a panel data set for colleges across differentyears. Statistics on male/female ratios, fraction of men/women in fraternities or sororities, policyvariables – such as a “safe house” for women on campus, as was started at MSU in 1994 – couldbe added as explanatory variables. The crime rate in the host town would be a good control.CARDSource: D. Card (1995), “Using Geographic Variation in College Proximity to Estimate theReturn to Schooling,” in Aspects of Labour Market Behavior: Essays in Honour of JohnVanderkamp. Ed. L.N. Christophides, E.K. Grant, and R. Swidinsky, 201-222. Toronto:University of Toronto Press.Professor Card kindly provided these data.Used in Text: pages 507-508, 527Notes: Computer Exercise C15.3 is important for analyzing these data. There, it is shown thatthe instrumental variable, nearc4, is actually correlated with IQ, at least for the subset of men forwhich an IQ score is reported. However, the correlation between nearc4 and IQ, once the otherexplanatory variables are netted out, is arguably zero. (At least, it is not statistically differentfrom zero.) In other words, nearc4 fails the exogeneity requirement in a simple linear model butit passes – at least using the crude test described above – if controls are added to the wageequation.For a more advanced course, a nice extension of Card’s analysis is to allow the return toeducation to differ by race. A relatively simple extension is to include blackeduc as anadditional explanatory variable; its natural instrument is blacknearc4.CATHOLICSource: Altonji, J.G., T.E. Elder, and C.R. Taber (2005), “An Evaluation of InstrumentalVariable Strategies for Estimating the Effects of Catholic Schooling,” Journal of HumanResources 40, 791-821.Professor Elder kindly provided a subset of the data, with some variables stripped away forconfidentiality reasons.Used in Text: pages 260-261, 530CEMENTSource: J. Shea (1993), “The Input-Output Approach to Instrument Selection,” Journal ofBusiness and Economic Statistics 11, 145-156.Professor Shea kindly provided these data.Used in Text: pages 5568Notes: Compared with Shea’s analysis, the producer price index (PPI) for fuels and power hasbeen replaced with the PPI for petroleum. The data are monthly and have not been seasonallyadjusted.CENSUS2000Source: Obtained from the United States Census Bureau by Professor Alberto Abadie of theHarvard Kennedy School of Government.Professor Abadie kindly provided the data.Used in Text: pages 485CEOSAL1Source: I took a random sample of data reported in the May 6, 1991 issue of Businessweek.Used in Text: pages 29, 32-33, 35, 154, 211, 252-253, 257, 662Notes: This kind of data collection is relatively easy for students just learning data analysis, andthe findings can be interesting. A good term project is to have students collect a similar data setusing a more recent issue of Businessweek, and to find additional variables that might explaindifferences in CEO compensation. My impression is that the public is still interested in CEOcompensation.An interesting question is whether the list of explanatory variables included in this data set nowexplain less of the variation in log(salary) than they used to.CEOSAL2Source: See CEOSAL1Used in Text: pages 62, 110, 154, 207, 324, 662Notes: Compared with CEOSAL1, in this CEO data set more information about the CEO, ratherthan about the company, is included.9CHARITYSource: P.H. Franses and R. Paap (2001), Quantitative Models in Marketing Research.Cambridge: Cambridge University Press.Professor Franses kindly provided the data.Used in Text: pages 63, 111, 260, 599Notes: This data set can be used to illustrate probit and Tobit models, and to study the linearapproximations to them.CONSUMPSource: I collected these data from the 1997 Economic Report of the President. Specifically, thedata come from Tables B71, B15, B29, and B32.Used in Text: pages 363-364, 391, 422, 548-549, 555, 640Notes: For a student interested in time series methods, updating this data set and using it in amanner similar to that in the text could be acceptable as a final project.CORNSource: G.E. Battese, R.M. Harter, and W.A. Fuller (1988), “An Error-Components Model forPrediction of County Crop Areas Using Survey and Satellite Data,” Journal of the AmericanStatistical Association 83, 28-36.This small data set is reported in the article.Used in Text: pages 745Notes: You could use these data to illustrate simple regression when the population interceptshould be zero: no corn pixels should predict no corn planted. The same can be done with thesoybean measures in the data set.COUNTYMURDERSSource: Compiled by J. Monroe Gamble for a Summer Research Opportunities Program (SROP)at Michigan State University, Summer 2014. Monroe obtained data from the U.S. CensusBureau, the FBI Uniform Crime Reports, and the Death Penalty Information Center.10Used in Text: pages 17, 64, 458, 490CPS78_85Source: Professor Henry Farber, now at Princeton University, compiled these data from the1978 and 1985 Current Population Surveys. Professor Farber kindly provided these data whenwe were colleagues at MIT.Used in Text: pages 427, 429-430, 454Notes: Obtaining more recent data from the CPS allows one to track, over a long period of time,the changes in the return to education, the gender gap, black-white wage differentials, and theunion wage premium.CPS91Source: Professor Daniel Hamermesh, at the University of Texas, compiled these data from theMay 1991 Current Population Survey. Professor Hamermesh kindly provided these data.Used in Text: page 599Notes: This is much bigger than the other CPS data sets even though the sample is restricted tomarried women. (CPS91 contains many more observations than MROZ, too.) In addition to theusual human capital variables for the women in the sample, we have information on the husband.Therefore, we can estimate a labor supply function as in Chapter 16, although the validity ofpotential experience as an IV for log(wage) is questionable. (MROZ contains an actualexperience variable.) Perhaps more convincing is to add hours to the wage offer equation, andinstrument hours with indicators for young and old children. This data set also contains a unionmembership indicator.The web site for the National Bureau of Economic Research makes it very easy now to downloadCPS data files in a variety of formats. Go to http://www.nber.org/data/cps_basic.html.CRIME1Source: J. Grogger (1991), “Certainty vs. Severity of Punishment,” Economic Inquiry 29, 297-309.Professor Grogger kindly provided a subset of the data he used in his article.Used in Text: pages 78, 169, 174, 243, 268, 291, 295-296, 581-582, 59711CRIME2Source: These data were collected by David Dicicco, a former MSU undergraduate, for a finalproject. They came from various issues of the County and City Data Book, and are for the years1982 and 1985. Unfortunately, I do not have the list of cities.Used in Text: pages 303-304, 439-440Notes: Very rich crime data sets, at the county, or even city, level, can be collected using theFBI’s Uniform Crime Reports. These data can be matched up with demographic and economicdata, at least for census years. The County and City Data Book contains a variety of statistics,but the years do not always match up. These data sets can be used investigate issues such as theeffects of casinos on city or county crime rates.CRIME3:Source: E. Eide (1994), Economics of Crime: Deterrence of the Rational Offender.Amsterdam: North Holland. The data come from Tables A3 and A6.Used in Text: pages 443-444, 455Notes: These data are for the years 1972 and 1978 for 53 police districts in Norway. Muchlarger data sets for more years can be obtained for the United States, although a measure of the“clear-up” rate is needed.CRIME4Source: From C. Cornwell and W. Trumball (1994), “Estimating the Economic Model of Crimewith Panel Data,” Review of Economics and Statistics 76, 360-366.Professor Cornwell kindly provided the data.Used in Text: pages 449-450, 456, 486, 556Notes: Computer Exercise C16.7 shows that variables that might seem to be good instrumentalvariable candidates are not always so good, especially after applying a transformation such asdifferencing across time. You could have the students do an IV analysis for just, say, 1987.DISCRIMSource: K. Graddy (1997), “Do Fast-Food Chains Price Discriminate on the Race and IncomeCharacteristics of an Area?” Journal of Business and Economic Statistics 15, 391-401.Professor Graddy kindly provided the data set.Used in Text: pages 111, 161, 663Notes: If you want to assign a common final project, this would be a good data set. There aremany possible dependent variables, namely, prices of various fast-food items. The key variable12is the fraction of the population that is black, along with controls for poverty, income, housingvalues, and so on. These data were also used in a famous study by David Card and Alan Kruegeron estimation of minimum wage effects on employment. See the book by Card and Krueger,Myth and Measurement, 1997, Princeton University Press, for a detailed analysis.DRIVINGSource: Freeman, D.G. (2007), “Drunk Driving Legislation and Traffic Fatalities: NewEvidence on BAC 08 Laws,” Contemporary Economic Policy 25, 293–308.Professor Freeman kindly provided the data.Used in Text: page 489; see also the general discussion on page 659Notes: Several more years of data are now available and may further shed light on theeffectiveness of several traffic laws.EARNSSource: Economic Report of the President, 1989, Table B47. The data are for the non-farmbusiness sector.Used in Text: pages 382, 390Notes: These data could be usefully updated, but changes in reporting conventions in morerecent ERPs may make that difficult.ECONMATHSource: Compiled by Professor Charles Ballard, Michigan State University Department ofEconomics.Professor Ballard kindly provided the data.Used in Text: 162, 177ELEM94_95Source: Culled from a panel data set used by Professor Leslie Papke in her paper “The Effects ofSpending on Test Pass Rates: Evidence from Michigan” (2005), Journal of Public Economics89, 821-839.Used in Text: pages 161, 330-331Notes: Starting in 1995, the Michigan Department of Education stopped reporting averageteacher benefits along with average salary. This data set includes both variables, at the schoollevel, and can be used to study the salary-benefits tradeoff, as in Chapter 4. There are a fewsuspicious benefits/salary ratios, and so this data set makes a good illustration of the impact ofoutliers in Chapter 9.13EXPENDSHARESSource: Blundell, R., A. Duncan, and K. Pendakur (1998), “Semiparametric Estimation andConsumer Demand,” Journal of Applied Econometrics 13, 435-461.I obtained these data from the Journal of Applied Econometrics data archive athttp://qed.econ.queensu.ca/jae/.Used in Text: pages 557-558Notes: The dependent variables in this data set – the expenditure shares – are necessarilybounded between zero and one. The linear model is at best an approximation, but the usual IVestimator likely gives good estimates of the average partial effects.ENGINSource: Thada Chaisawangwong, a former graduate student at MSU, obtained these data for aterm project in applied econometrics. They come from the Material Requirement PlanningSurvey carried out in Thailand during 1998.Used in Text: not usedNotes: This is a nice change of pace from wage data sets for the United States. These data arefor engineers in Thailand, and represents a more homogeneous group than data sets that consistof people across a variety of occupations. Plus, the starting salary is also provided in the data set,so factors affecting wage growth – and not just wage levels at a given point in time – can bestudied. This is a good data set for a common term project that tests basic understanding ofmultiple regression and the interpretation of models with a logarithm for a dependent variable.EZANDERSSource: L.E. Papke (1994), “Tax Policy and Urban Development: Evidence from the IndianaEnterprise Zone Program,” Journal of Public Economics 54, 37-49.Professor Papke kindly provided these data.Used in Text: page 363Notes: These are actually monthly unemployment claims for the Anderson enterprise zone.Papke used annualized data, across many zones and non-zones, in her original analysis.EZUNEMSource: See EZANDERSUsed in Text: pages 449, 486-48714Notes: A very good project is to have students analyze enterprise, empowerment, renaissance,or opportunity zone policies in their home states. Many states now have such programs. Or,there are also national programs. A few years of panel data straddling periods of zonedesignation, at the city or zip code level, could make a nice study.FAIRSource: R.C. Fair (1996), “Econometrics and Presidential Elections,” Journal of EconomicPerspectives 10, 89-102.The data set is provided in the article.Used in Text: pages 350-351, 420, 422Notes: An updated version of this data set, through the 2004 election, is available at ProfessorFair’s web site at Yale University: http://fairmodel.econ.yale.edu/rayfair/pdf/2001b.htm.Students might want to try their own hands at predicting the most recent election outcome, butthey should be restricted to no more than a handful of explanatory variables because of the smallsample size.FERTIL1Source: W. Sander, “The Effect of Women’s Schooling on Fertility,” Economics Letters 40,229-233.Professor Sander kindly provided the data, which are a subset of what he used in his article. Hecompiled the data from various years of the National Opinion Resource Center’s General SocialSurvey.Used in Text: pages 428-429, 453, 521, 597, 646Notes: (1) Much more recent data can be obtained from the National Opinion Research Centerwebsite, http://www.norc.org/GSS+Website/Download/. Very rich pooled cross sections can beconstructed to study a variety of issues – not just changes in fertility over time.(2) It would be interesting to analyze a similar data set for a developing country, especiallywhere efforts have been made to emphasize birth control. Some measure of access to birthcontrol could be useful if it varied by region. Sometimes, one can find policy changes in theadvertisement or availability of contraceptives.FERTIL2Source: These data were obtained by James Heakins, a former MSU undergraduate, for a termproject. They come from Botswana’s 1988 Demographic and Health Survey.Used in Text: page 526-52715Notes: Currently, this data set is used only in one computer exercise. Since the dependentvariable of interest – number of living children or number of children every born – is a countvariable, the Poisson regression model discussed in Chapter 17 can be used. However, somecare is required to combine Poisson regression with an endogenous explanatory variable (educ).I refer you to Chapter 19 of my book Econometric Analysis of Cross Section and Panel Data.Even in the context of linear models, much can be done beyond Computer Exercise C15.2. At aminimum, the binary indicators for various religions can be added as controls. One might alsointeract the schooling variable, educ, with some of the exogenous explanatory variables.FERTIL3Source: L.A. Whittington, J. Alm, and H.E. Peters (1990), “Fertility and the PersonalExemption: Implicit Pronatalist Policy in the United States,” American Economic Review 80,545-556.The data are given in the article.Used in Text: pages 346-347, 355, 362, 363, 363, 381, 384-385, 421, 634, 639-640FISHSource: K Graddy (1995), “Testing for Imperfect Competition at the Fulton Fish Market,”RAND Journal of Economics 26, 75-92.Professor Graddy’s collaborator on a later paper, Professor Joshua Angrist at MIT, kindlyprovided me with these data.Used in Text: pages 422-423, 556-557Notes: This is a nice example of how to go about finding exogenous variables to use asinstrumental variables. Often, weather conditions can be assumed to affect supply while havinga negligible effect on demand. If so, the weather variables are valid instrumental variables forprice in the demand equation. It is a simple matter to test whether prices vary with weatherconditions by estimating the reduced form for price.FRINGESource: F. Vella (1993), “A Simple Estimator for Simultaneous Models with CensoredEndogenous Regressors,” International Economic Review 34, 441-457.Professor Vella kindly provided the data.Used in Text: page 596-597Notes: Currently, this data set is used in only one Computer Exercise – to illustrate the Tobitmodel. It can be used much earlier. First, one could just ignore the pileup at zero and use alinear model where any of the hourly benefit measures is the dependent variable. Anotherpossibility is to use this data set for a problem set in Chapter 4, after students have read Example4.10. That example, which uses teacher salary/benefit data at the school level, finds the expected16tradeoff, although it appears to less than one-to-one. By contrast, if you do a similar analysiswith FRINGE, you will not find a tradeoff. A positive coefficient on the benefit/salary ratio isnot too surprising because we probably cannot control for enough factors, especially whenlooking across different occupations. The Michigan school-level data is more aggregated thanone would like, but it does restrict attention to a more homogeneous group: high school teachersin Michigan.GPA1Source: Christopher Lemmon, a former MSU undergraduate, collected these data from a surveyhe took of MSU students in Fall 1994.Used in Text: pages 72, 74, 77, 113. 127. 155. 162. 225. 256. 286. 291Notes: This is a nice example of how students can obtain an original data set by focusing locallyand carefully composing a survey.GPA2Source: For confidentiality reasons, I cannot provide the source of these data. I can say that theycome from a midsize research university that also supports men’s and women’s athletics at theDivision I level.Used in Text: pages 104, 178, 202-203, 204-205, 215, 252, 256-257GPA3Source: See GPA2Used in Text: pages 237-238, 266-267, 287-288, 444, 455HAPPINESSSource: Subset of data collected by Kevin Williams for a McNair Scholars project in Summer2008 at Michigan State University. The data come from several waves of the General SocialSurvey, and is therefore a pooled cross sectional data set. Professor Williams, now at YaleUniversity, kindly provided the data.Used in Text: not usedNotes: This data set can be used to estimate models of self-reported “happiness,” includingstudying whether the effects of certain variables – such as education, gender, race, and havingchildren –changed in importance from the mid-1990s to the mid-2000s. For a similar example,see how FERTIL1 is used in Example 13.1 in the text.HPRICE1Source: Collected from the real estate pages of the Boston Globe during 1990. These homessold in the Boston, MA area.17Used in Text: pages 109-110, 148-149, 155-156, 160, 205, 216, 226-227, 270-271, 273, 290,297-298Notes: Typically, it is very easy to obtain data on selling prices and characteristics of homes,using publicly available databases. It is interesting to match the information on houses withother information – such as local crime rates, quality of the local schools, pollution levels, and soon – and estimate the effects of such variables on housing prices.HPRICE2Source: D. Harrison and D.L. Rubinfeld (1978), “Hedonic Housing Prices and the Demand forClean Air,” by Harrison, D. and D.L.Rubinfeld, Journal of Environmental Economics andManagement 5, 81-102.Diego Garcia, a former Ph.D. student in economics at MIT, kindly provided these data, which heobtained from the book Regression Diagnostics: Identifying Influential Data and Sources ofCollinearity, by D.A. Belsey, E. Kuh, and R. Welsch, 1990. New York: Wiley.Used in Text: pages 106, 130, 185, 186-187, 190-191Notes: The census contains rich information on variables such as median housing prices,median income levels, average family size, and so on, for fairly small geographical areas. Ifsuch data can be merged with pollution data, one can update the Harrison and Rubinfeld study.Presumably, this has been done in academic journals.HSEINVSource: D. McFadden (1994), “Demographics, the Housing Market, and the Welfare of theElderly,” in D.A. Wise (ed.), Studies in the Economics of Aging. Chicago: University ofChicago Press, 225-285.The data are contained in the article.Used in Text: pages 354-355, 358, 390, 609-610, 638, 783HTVSource: J.J. Heckman, J.L. Tobias, and E. Vytlacil (2003), “Simple Estimators for TreatmentParameters in a Latent-Variable Framework,” Review of Economics and Statistics 85, 748-755.Professor Tobias kindly provided the data, which were obtained from the 1991 NationalLongitudinal Survey of Youth. All people in the sample are males age 26 to 34. Forconfidentiality reasons, I have included only a subset of the variables used by the authors.Used in Text: pages 529, 599-600Notes: Because an ability measure is included in this data set, it can be used as anotherillustration of including proxy variables in regression models. See Chapter 9. Also, one can trythe IV procedure with the ability measure included as an exogenous explanatory variable.18INFMRTSource: Statistical Abstract of the United States, 1990 and 1994. (For example, the infantmortality rates come from Table 113 in 1990 and Table 123 in 1994.)Used in Text: pages 320-321, 328Notes: An interesting exercise is to add the percentage of the population on AFDC (afdcper) tothe infant mortality equation. Pooled OLS and first differencing can give very differentestimates. Adding the years 1998 and 2002 and applying fixed effects seems natural.Intervening years can be added, too, although variation in the key variables from year to yearmight be minimal.INJURYSource: B.D. Meyer, W.K. Viscusi, and D.L. Durbin (1995), “Workers’ Compensation andInjury Duration: Evidence from a Natural Experiment,” American Economic Review 85, 322-340.Professor Meyer kindly provided the data.Used in Text: pages 435-436, 453Notes: This data set also can be used to illustrate the Chow test in Chapter 7. In particular,students can test whether the regression functions differ between Kentucky and Michigan. Or,allowing for different intercepts for the two states, do the slopes differ? A good lesson from thisexample is that a small R-squared is compatible with the ability to estimate the effects of apolicy. Of course, for the Michigan data, which has a smaller sample size, the estimated effect ismuch less precise (but of virtually identical magnitude).INTDEFSource: Economic Report of the President, 2004, Tables B64, B73, and B79.Used in Text: pages 345, 363, 415, 527INTQRTSource: From Salomon Brothers, Analytical Record of Yields and Yield Spreads, 1990. Thefolks at Salomon Brothers kindly provided the Record at no charge when I was an assistantprofessor at MIT.Used in Text: pages 388-389, 612, 617, 620, 621, 639, 640Notes: A nice feature of the Salomon Brothers data is that the interest rates are not averagedover a month or quarter – they are end-of-month or end-of-quarter rates. Asset pricing theoriesapply to such “point-sampled” data, and not to averages over a period. Most other sources reportmonthly or quarterly averages. This is a good data set to update and test whether current data aremore or less supportive of basic asset pricing theories.19INVENSource: Economic Report of the President, 1997, Tables B4, B20, B61, and B71.Used in Text: pages 391, 421-422, 423, 614, 783JTRAINSource: H. Holzer, R. Block, M. Cheatham, and J. Knott (1993), “Are Training SubsidiesEffective? The Michigan Experience,” Industrial and Labor Relations Review 46, 625-636.The authors kindly provided the data.Used in Text: pages 133-134, 156, 226, 244-245, 328, 444-446, 456, 464-465, 469, 486, 521-522, 730-731, 740-741, 742JTRAIN2Source: R.J. Lalonde (1986), “Evaluating the Econometric Evaluations of Training Programswith Experimental Data,” American Economic Review 76, 604-620.Professor Jeff Biddle, at MSU, kindly passed the data set along to me. He obtained it fromProfessor Lalonde.Used in Text: pages 16, 329-330Notes: Professor Lalonde obtained the data from the National Supported Work Demonstrationjob-training program conducted by the Manpower Demonstration Research Corporation in themid 1970s. Training status was randomly assigned, so this is essentially experimental data.Computer Exercise C17.8 looks only at the effects of training on subsequent unemploymentprobabilities. For illustrating the more advanced methods in Chapter 17, a good exercise wouldbe to have the students estimate a Tobit of re78 on train, and obtain estimates of the expectedvalues for those with and without training. These can be compared with the sample averages.JTRAIN3Source: R.H. Dehejia and S. Wahba (1999), “Causal Effects in Nonexperimental Studies:Reevaluating the Evaluation of Training Programs,” Journal of the American StatisticalAssociation 94, 1053-1062.Professor Sergio Firpo, at Insper Institute of Education and Research inSão Paulo, has used thisdata set in his work. He kindly provided it to me. This data set is a subset of that originally usedby Lalonde in the study cited for JTRAIN2.Used in Text: pages 329-330, 45720JTRAIN98Source: This is a data set I created many years ago intended as an update to the files JTRAIN2and JTRAIN3. While the data were partly generated by me, the data attributes are similar to datasets used to evaluate job training programs.Used in Text: 101-102, 248, 601Notes: The response variables, earn98 and unem98, both have discreteness: the former is acorner solutions (takes on the value zero and then a range of strictly positive values) and thelatter is binary. One could use these in an exercise using methods in Chapter 17. unem98 can beused in a probit or logit model, earn98 in a Tobit model, or in Poisson regression (withoutassuming, of course, that the Poisson distribution is correct).KIELMCSource: K.A. Kiel and K.T. McClain (1995), “House Prices During Siting Decision Stages: TheCase of an Incinerator from Rumor through Operation,” Journal of Environmental Economicsand Management 28, 241-255.Professors Kiel and McClain kindly provided the data, of which I used only a subset.Used in Text: pages 214, 431-434, 452, 454LABSUPSource: The subset of data for black or Hispanic women used in J.A. Angrist and W.E. Evans(1998), “Used in Text: pages 530-531Notes: This example can promote an interesting discussion of instrument validity, and inparticular, how a variable that is beyond our control – for example, whether the first two childrenhave the same gender – can, nevertheless, affect subsequent economic choices. Students areasked to think about such issues in Computer Exercise C13 in Chapter 15. A more egregiousversion of this mistake would be to treat a variable such as age as a suitable instrument because itis beyond our control: clearly age has a direct effect on many economic outcomes that wouldplay the role of the dependent variable.LAWSCH85Source: Collected by Kelly Barnett, an MSU economics student, for use in a term project. Thedata come from two sources: The Official Guide to U.S. Law Schools, 1986, Law SchoolAdmission Services, and The Gourman Report: A Ranking of Graduate and ProfessionalPrograms in American and International Universities, 1995, Washington, D.C.Used in Text: pages 105, 108, 159-160, 231-23221Notes: More recent versions of both cited documents are available. One could try a similaranalysis for, say, MBA programs or Ph.D. programs in economics. Quality of placements maybe a good dependent variable, and measures of business school or graduate program qualitycould be included among the explanatory variables. Of course, one would want to control forfactors describing the incoming class so as to isolate the effect of the program itself.LOANAPPSource: W.C. Hunter and M.B. Walker (1996), “The Cultural Affinity Hypothesis andMortgage Lending Decisions,” Journal of Real Estate Finance and Economics 13, 57-70.Professor Walker kindly provided the data.Used in Text: pages 257-258, 291, 329, 596Notes: These data were originally used in a famous study by researchers at the Boston FederalReserve Bank. See A. Munnell, G.M.B. Tootell, L.E. Browne, and J. McEneaney (1996),“Mortgage Lending in Boston: Interpreting HMDA Data,” American Economic Review 86, 25-53.LOWBRTHSource: Source: Statistical Abstract of the United States, 1990, 1993, and 1994.Used in Text: not usedNotes: This data set can be used very much like INFMRT. It contains two years of state-levelpanel data. In fact, it is a superset of INFMRT. The key is that it contains information on lowbirth weights, as well as infant mortality. It also contains state identifies, so that several years ofmore recent data could be added for a term project. Putting in the variable afcdprc and its squareleads to some interesting findings for pooled OLS and fixed effects (first differencing). Afterdifferencing, you can even try using the change in the AFDC payments variable as aninstrumental variable for the change in afdcprc.MATHPNLSource: Dr. Leslie Papke, an economics professor at MSU, collected these data from MichiganDepartment of Education web site, www.michigan.gov/mde. These are district-level data, whichProfessor Papke kindly provided. She has used building-level data in “The Effects of Spendingon Test Pass Rates: Evidence from Michigan” (2005), Journal of Public Economics 89, 821-839.Used in Text: pages 456, 487-488MEAP00Source: Michigan Department of Education, www.michigan.gov/mdeUsed in Text: pages 218, 29222MEAP01Source: Michigan Department of Education, www.michigan.gov/mdeUsed in Text: page 16Notes: This is another good data set to compare simple and multiple regression estimates. Theexpenditure variable (in logs, say) and the poverty measure (lunch) are negatively correlated inthis data set. A simple regression of math4 on lexppp gives a negative coefficient. Controllingfor lunch makes the spending coefficient positive and significant.MEAP93Source: I collected these data from the old Michigan Department of Education web site. SeeMATHPNL for the current web site. I used data on most high schools in the state of Michiganfor 1993. I dropped some high schools that had suspicious-looking data.Used in Text: pages 44-45, 63, 110-111, 125-126, 149-150, 158, 212, 325, 329, 329, 660Notes: Many states have data, at either the district or building level, on student performance andspending. A good exercise in data collection and cleaning is to have students find such data for aparticular state, and to put it into a form that can be used for econometric analysis.MEAPSINGLESource: Collected by Professor Leslie Papke, an economics professor at MSU, from theMichigan Department of Education web site, www.michigan.gov/mde, and the U.S. CensusBureau. Professor Papke kindly provided the data.Used in Text: 110-111, 158-159, 213MINWAGESource: P. Wolfson and D. Belman (2004), “The Minimum Wage: Consequences for Prices andQuantities in Low-Wage Labor Markets,” Journal of Business & Economic Statistics 22, 296-311.Professor Belman kindly provided the data.Used in Text: pages 365, 393, 424, 641Notes: The sectors corresponding to the different numbers in the data file are provided in theWolfson and Bellman and article.23MLB1Source: Collected by G. Mark Holmes, a former MSU undergraduate, for a term project. Thesalary data were obtained from the New York Times, April 11, 1993. The baseball statistics arefrom The Baseball Encyclopedia, 9th edition, and the city population figures are from theStatistical Abstract of the United States.Used in Text: pages 140-143, 160, 229, 235-236, 256-257Notes: The baseball statistics are career statistics through the 1992 season. Players whose raceor ethnicity could not be easily determined were not included. It should not be too difficult toobtain the city population and racial composition numbers for Montreal and Toronto for 1993.Of course, the data can be pretty easily obtained for more recent players.MROZSource: T.A. Mroz (1987), “The Sensitivity of an Empirical Model of Married Women’s Hoursof Work to Economic and Statistical Assumptions,” Econometrica 55, 765-799.Professor Ernst R. Berndt, of MIT, kindly provided the data, which he obtained from ProfessorMroz.Used in Text: pages 240, 253, 285, 501, 511, 516, 518, 543-544, 555, 568-569, 575-576, 591-592, 597MURDERSource: From the Statistical Abstract of the United States, 1995 (Tables 310 and 357), 1992(Table 289). The execution data originally come from the U.S. Bureau of Justice Statistics,Capital Punishment Annual.Used in Text: pages 457, 487, 527-528Notes: The data set COUNTYMURDERS includes information on executions and murder ratesat the county level, and provides more variation.NBASALSource: Collected by Christopher Torrente, a former MSU undergraduate, for a term project.He obtained the salary data and the career statistics from The Complete Handbook of ProBasketball, 1995, edited by Zander Hollander. New York: Signet. The demographic information(marital status, number of children, and so on) was obtained from the teams’ 1994-1995 mediaguides.Used in Text: pages 216-217, 258Notes: A panel version of this data set could be useful for further isolating productivity effectsof marital status. One would need to obtain information on enough different players in at least24two years, where some players who were not married in the initial year are married in later years.Fixed effects (or first differencing, for two years) is the natural estimation method.NCAA_RPISource: Data on NCAA men’s basketball teams, collected by Weizhao Sun for a senior seminarproject in sports economics at Michigan State University, Spring 2017. He used various sources,including www.espn.com and www.teamrankings.com/ncaa-basketball/rpi-ranking/rpi-rating-byteam.Used in Text: not usedNotes: This is a nice example of how multiple regression analysis can be used to determinewhether rankings compiled by experts – the so-called pre-season RPI in this case – provideadditional information beyond what we can obtain from widely available data bases. A simpleand interesting question is whether, once the previous year’s post-season RPI is controlled for,does the pre-season RPI – which is supposed to add information on recruiting and playerdevelopment – help to predict performance (such as win percentage or making it to the NCAAmen’s basketball tournament). For the binary outcome that indicates making it to the NCAAtournament, a probit or logit model can be used for courses that introduce more advancedmethods. There are some other interesting variables, such as coaching experience, that can beincluded, too.NYSESource: These are Wednesday closing prices of value-weighted NYSE average, available inmany publications. I do not recall the particular source I used when I collected these data atMIT. Probably the easiest way to get similar data is to go to the NYSE web site, www.nyse.com.Used in Text: pages 352-353, 368, 393, 398, 399, 595OKUNSource: Economic Report of the President, 2007, Tables B4 and B42.Used in Text: 392, 423-424OPENNESSSource: D. Romer (1993), “Openness and Inflation: Theory and Evidence,” Quarterly Journalof Economics 108, 869-903.The data are included in the article.Used in Text: pages 544-545, 55525PENSIONSource: L.E. Papke (2004), “Individual Financial Decisions in Retirement Saving: The Role ofParticipant-Direction,” Journal of Public Economics 88, 39-61.Professor Papke kindly provided the data. She collected them from the National LongitudinalSurvey of Mature Women, 1991.Used in Text: page 488PHILLIPSSource: Economic Report of the President, 2004, Tables B42 and B64.Used in Text: pages 344-345, 364-365, 375, 390-391, 392, 392, 403, 404, 412, 423, 528, 613,625, 628, 639, 770PNTSPRDSource: Collected by Scott Resnick, a former MSU undergraduate, from various newspapersources.Used in Text: pages 271, 560, 623Notes: The data are for the 1994-1995 men’s college basketball seasons. The spread is for theday before the game was played. One might collect more recent data and determine whether thespread has become a less accurate predictor of the actual outcome in more recent years. In otherwords, in the simple regression of the actual score differential on the spread, is the variancelarger in more recent years. (We should fully expect the slope coefficient not to be statisticallydifferent from one.)PRISONSource: S.D. Levitt (1996), “The Effect of Prison Population Size on Crime Rates: Evidencefrom Prison Overcrowding Legislation,” Quarterly Journal of Economics 111, 319-351.Professor Levitt kindly provided me with the data, of which I used a subset.Used in Text: pages 551PRMINWGESource: A.J. Castillo-Freeman and R.B. Freeman (1992), “When the Minimum Wage ReallyBites: The Effect of the U.S.-Level Minimum Wage on Puerto Rico,” in Immigration and theWork Force, edited by G.J. Borjas and R.B. Freeman, 177-211. Chicago: University of ChicagoPress.The data are reported in the article.26Used in Text: pages 345-346, 356-357, 400, 405Notes: Given the ongoing debate on the employment effects of the minimum wage, this wouldbe a great data set to try to update. The coverage rates are the most difficult variables toconstruct.RECIDSource: C.-F. Chung, P. Schmidt, and A.D. Witte (1991), “Survival Analysis: A Survey,”Journal of Quantitative Criminology 7, 59-98.Professor Chung kindly provided the data.Used in Text: pages 584-585, 597RDCHEMSource: From Businessweek R&D Scoreboard, October 25, 1991.Used in Text: pages 63, 135-136, 154-155, 198, 211, 317-318, 328-329Notes: It would be interesting to collect more recent data and see whether the R&D/firm sizerelationship has changed over time.RDTELECSource: See RDCHEMUsed in Text: not usedNotes: According to these data, the R&D/firm size relationship is different in thetelecommunications industry than in the chemical industry: there is pretty strong evidence thatR&D intensity decreases with firm size in telecommunications. Of course, that was in 1991. Thedata could easily be updated, and a panel data set could be constructed for more advancedcourses.RENTALSource: David Harvey, a former MSU undergraduate, collected the data for 64 “college towns”from the 1980 and 1990 United States censuses.Used in Text: pages 155, 444, 454-455, 486Notes: These data can be used in a somewhat crude simultaneous equations analysis, eitherfocusing on one year or pooling the two years. (In the latter case, in an advanced class, youmight have students compute the standard errors robust to serial correlation across the two timeperiods.) The demand equation would have ltothsg as a function of lrent, lavginc, and lpop. Thesupply equation would have ltothsg as a function of lrent, pctst, and lpop. Thus, in estimatingthe demand function, pctstu is used as an IV for lrent. Clearly one can quibble with excluding27pctstu from the demand equation, but the estimated demand function gives a negative priceeffect.Getting information for 2000 and 2010, and adding many more college towns, would make for amuch better analysis. Information on number of spaces in on-campus dormitories would be a bigimprovement, too.RETURNSource: Collected by Stephanie Balys, a former MSU undergraduate, from the New York StockExchange and Compustat.Used in Text: page 157Notes: More can be done with this data set. Recently, I discovered that lsp90 does appear topredict return (and the log of the 1990 stock price works better than sp90). I am a littlesuspicious, but you could use the negative coefficient on lsp90 to illustrate “reversion to themean.”SAVINGSource: UnknownUsed in Text: not usedNotes: I remember entering this data set in the late 1980s, and I am pretty sure it came directlyfrom an introductory econometrics text. But so far my search has been fruitless. If anyone runsacross this data set, I would appreciate knowing about it.SCHOOL93_98Source: L.E. Papke (2005), “The Effects of Spending on Test Pass Rates: Evidence fromMichigan,” Journal of Public Economics 89, 821-839.Used in Text: page 491Notes: This is closer to the data actually used in the Papke paper as it is at the school (building)level. It is unbalanced because data on scores and some of the spending and other variables ismissing for some schools. While the usual RE and FE methods can be applied directly, obtainingthe correlated random effects version of the Hausman test is more advance. Computer Exercise17 in Chapter 14 walks the reader through it.SLEEP75Source: J.E. Biddle and D.S. Hamermesh (1990), “Sleep and the Allocation of Time,” Journalof Political Economy 98, 922-943.Professor Biddle kindly provided the data.28Used in Text: pages 62, 105, 156-157, 251, 257, 290Notes: In their article, Biddle and Hamermesh include an hourly wage measure in the sleepequation. An econometric problem that arises is that the hourly wage is missing for those whodo not work. Plus, the wage offer may be endogenous (even if it were always observed). Biddleand Hamermesh employ extensions of the sample selection methods in Section 17.5. See theirarticle for details.SLP75_81Source: See SLEEP75Used in Text: pages 442-443SMOKESource: J. Mullahy (1997), “Instrumental-Variable Estimation of Count Data Models:Applications to Models of Cigarette Smoking Behavior,” Review of Economics and Statistics 79,596-593.Professor Mullahy kindly provided the data.Used in Text: pages 177, 280-281, 288, 291-292, 555, 598-599Notes: If you want to do a “fancy” IV version of Computer Exercise C16.1, you could estimatea reduced form count model for cigs using the Poisson regression methods in Section 17.3, andthen use the fitted values as an IV for cigs. Presumably, this would be for a fairly advancedclass.TRAFFIC1Source: I collected these data from two sources, the 1992 Statistical Abstract of the UnitedStates (Tables 1009, 1012) and A Digest of State Alcohol-Highway Safety Related Legislation,1985 and 1990, published by the U.S. National Highway Traffic Safety Administration.Used in Text: pages 444, 446Notes: In addition to adding recent years, this data set could really use state-level tax rates onalcohol. Other important law changes include defining driving under the influence as having ablood alcohol level of .08 or more, which many states have adopted since the 1980s. The trendreally picked up in the 1990s and continued through the 2000s. The data set DRIVING is morecomplete and more recent, but it is also more complicated.TRAFFIC2Source: P.S. McCarthy (1994), “Relaxed Speed Limits and Highway Safety: New Evidencefrom California,” Economics Letters 46, 173-179.Professor McCarthy kindly provided the data.29Used in Text: pages 364, 392, 422, 641, 659Notes: Many states have changed maximum speed limits and imposed seat belt laws over thepast 25 years. Data similar to those in TRAFFIC2 should be fairly easy to obtain for a particularstate. One should combine this information with changes in a state’s blood alcohol limit and thepassage of per se and open container laws.TWOYEARSource: T.J. Kane and C.E. Rouse (1995), “Labor-Market Returns to Two- and Four-YearColleges,” American Economic Review 85, 600-614.With Professor Rouse’s kind assistance, I obtained the data from her web site at PrincetonUniversity.Used in Text: pages 137-139, 160, 254, 329Notes: As possible extensions, students can explore whether the returns to two-year or four-yearcolleges depend on race or gender. This is partly done in Problem 7.9 but where college isaggregated into one number. Also, should experience appear as a quadratic in the wagespecification?VOLATSource: J.D. Hamilton and L. Gang (1996), “Stock Market Volatility and the Business Cycle,”Journal of Applied Econometrics 11, 573-593.I obtained these data from the Journal of Applied Econometrics data archive athttp://qed.econ.queensu.ca/jae/Used in Text: pages 364, 637-640VOTE1Source: M. Barone and G. Ujifusa, The Almanac of American Politics, 1992. Washington, DC:National Journal.Used in Text: pages 31, 36, 159-160, 215-216, 290, 663VOTE2Source: See VOTE1Used in Text: pages 324-325, 444, 455-456, 663Notes: These are panel data, at the Congressional district level, collected for the 1988 and 1990U.S. House of Representative elections. Of course, much more recent data are available, possiblyeven in electronic form.30VOUCHERSource: Rouse, C.E. (1998), “Private School Vouchers and Student Achievement: AnEvaluation of the Milwaukee Parental Choice Program,” Quarterly Journal of Economics 113,553-602.Professor Rouse kindly provided the original data set from her paper.Used in Text: pages 529-530Notes: This is a condensed version of the data set used by Professor Rouse. The original data sethad missing information on many variables, including pre-program and post-program test scores.I did not impute any missing data and have dropped observations that were unusable withoutfilling in missing data. There are 990 students in the current data set but pre-program test scoresare available for only 328 of them.This is a good example of where eligibility for a program is randomized but participation neednot be. In addition, even if we look at just the effect of eligibility (captured in the variableselectyrs) on the math test score (mnce), we need to confront the fact that attrition (studentsleaving the district) can bias the results. Controlling for the pre-policy test score, mnce90, canhelp – but at the cost of losing two-thirds of the observations. A simple regression of mnce onselectyrs followed by a multiple regression that adds mnce90 as a control is informative.The selectyrs dummy variables can be used as instrumental variables for the choiceyrs variableto try to estimate the effect of actually participating in the program (rather than estimating the socalled intention-to-treat effect). Computer Exercise C15.11 steps through the details.WAGE1Source: These are data from the 1976 Current Population Survey, collected by Henry Farberwhen he and I were colleagues at MIT in 1988.Used in Text: pages 6, 30-31, 33, 73, 86, 123, 178, 189, 214, 222, 224, 227, 228-229, 232-233,235, 257, 265-266, 316, 644Notes: Barry Murphy, of the University of Portsmouth in the UK, has pointed out that forseveral observations the values for exper and tenure are in logical conflict. In particular, forsome workers the number of years with current employer (tenure) is greater than overall workexperience (exper). At least some of these conflicts are due to the definition of exper as“potential” work experience, but probably not all. Nevertheless, I am using the data set as it wassupplied to me.WAGE2Source: M. Blackburn and D. Neumark (1992), “Unobserved Ability, Efficiency Wages, andInterindustry Wage Differentials,” Quarterly Journal of Economics 107, 1421-1436.Professor Neumark kindly provided the data, of which I used just the data for 1980.31Used in Text: pages 63, 104-105, 110, 160, 212, 215, 256, 301-302, 328, 502, 515, 526, 528-529, 644Notes: As with WAGE1, there are some clear inconsistencies among the variables tenure, exper,and age. I have not been able to track down the source of the inconsistency, and so any changeswould be effectively arbitrary. Instead, I am using the data as provided by the authors of theabove QJE article.WAGEPANSource: F. Vella and M. Verbeek (1998), “Whose Wages Do Unions Raise? A Dynamic Modelof Unionism and Wage Rate Determination for Young Men,” Journal of Applied Econometrics13, 163-183.I obtained the data from the Journal of Applied Econometrics data archive athttp://qed.econ.queensu.ca/jae/. The JAE data archive is generally a nice resource forundergraduates looking to replicate or extend a published study.Used in Text: pages 457, 465, 472WAGEPRCSource: Economic Report of the President, various years.Used in Text: pages 388, 421, 638Notes: These monthly data run from January 1964 through October 1987. The consumer priceindex averages to 100 in 1967. An updated set of data can be obtained electronically fromhttp://www.gpo.gov/fdsys/browse/collection.action?collectionCode=ERP.WINESource: These data were reported in a New York Times article, December 28, 1994.Used in Text: not usedNotes: The dependent variables deaths, heart, and liver each can be regressed on alcohol as nicesimple regression examples. The conventional wisdom is that wine is good for the heart but notfor the liver, something that is apparent in the regressions. Because the number of observations issmall, this can be a good data set to illustrate calculation of the OLS estimates and statistics.
Get professional assignment help cheaply
Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?
Whichever your reason may is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.
Our essay writers are graduates with diplomas, bachelor, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college diploma. When assigning your order, we match the paper subject with the area of specialization of the writer.
Why choose our academic writing service?
Plagiarism free papers
Timely delivery
Any deadline
Skilled, Experienced Native English Writers
Subject-relevant academic writer
Adherence to paper instructions
Ability to tackle bulk assignments
Reasonable prices
24/7 Customer Support
Get superb grades consistently
Get Professional Assignment Help Cheaply
Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?
Whichever your reason may is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.
Our essay writers are graduates with diplomas, bachelor’s, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college diploma. When assigning your order, we match the paper subject with the area of specialization of the writer.
Why Choose Our Academic Writing Service?
Plagiarism free papers
Timely delivery
Any deadline
Skilled, Experienced Native English Writers
Subject-relevant academic writer
Adherence to paper instructions
Ability to tackle bulk assignments
Reasonable prices
24/7 Customer Support
Get superb grades consistently
How It Works
1. Place an order
You fill all the paper instructions in the order form. Make sure you include all the helpful materials so that our academic writers can deliver the perfect paper. It will also help to eliminate unnecessary revisions.
2. Pay for the order
Proceed to pay for the paper so that it can be assigned to one of our expert academic writers. The paper subject is matched with the writer’s area of specialization.
3. Track the progress
You communicate with the writer and know about the progress of the paper. The client can ask the writer for drafts of the paper. The client can upload extra material and include additional instructions from the lecturer. Receive a paper.
4. Download the paper
The paper is sent to your email and uploaded to your personal account. You also get a plagiarism report attached to your paper.
PLACE THIS ORDER OR A SIMILAR ORDER WITH Essay fount TODAY AND GET AN AMAZING DISCOUNT
The post essay on program inputs appeared first on Essay fount.
What Students Are Saying About Us
.......... Customer ID: 12*** | Rating: ⭐⭐⭐⭐⭐"Honestly, I was afraid to send my paper to you, but you proved you are a trustworthy service. My essay was done in less than a day, and I received a brilliant piece. I didn’t even believe it was my essay at first 🙂 Great job, thank you!"
.......... Customer ID: 11***| Rating: ⭐⭐⭐⭐⭐
"This company is the best there is. They saved me so many times, I cannot even keep count. Now I recommend it to all my friends, and none of them have complained about it. The writers here are excellent."
"Order a custom Paper on Similar Assignment at essayfount.com! No Plagiarism! Enjoy 20% Discount!"
