Articles omitted from the 1980s survey.

1

Jacobs, Rodney L., and Robert A. Jones

"Price Expectations in the United States: 1947-75"

1980(3)

yes

"The estimation entails finding the values of the adaptation parameters and initial conditions which minimizes the sum of transformed residuals . . . . We use the technique of Donald Marquardt" (p. 274). "The F-test cannot reject the hypothesis of zero drift at the 95 percent level of significance" (p. 274).

2

Santomero, Anthony M., and John H. Seater

"Partial Adjustment in the Demand for Money: Theory and Empirics"

1981(4)

yes

See equation (21), p. 573. "Estimation is by the Cochrane-Orcutt technique; OLS results suggest severe autocorrelation of the residuals." Equation (23), p. 575 also. "All variables are significant at the 1 percent level . . . " (p. 575).

3

Cooley, Thomas F., and Stephen F. LeRoy

"Identification and Estimation of Money Demand"

1981(5)

yes

"Least squares applied to the basic equation . . . yields the following estimates (standard errors in parentheses)" (p. 834).

4

Kotlikoff, Laurence J., Avia Spivak, and Lawrence H. Summers

"The Adequacy of Savings"

1982(5)

yes

"Table 6 presents regression . . . " (p. 1066) with coefficient estimates and t statistics in parentheses. "The three variables reflecting the form and timing of lifetime income are all highly significant in all or our regressions" (p. 1067).

5

Leamer, Edward E.

"Let's Take the Con Out of Econometrics"

1983(1)

yes

Regression example based on state-by-state murder rates in 1950 appears on pp. 40-42.

6

Pindyck, Robert S.

"Risk, Inflation, and the Stock Market

1984(3)

yes

Although regression and statistical significance does not appear prominently, it is there. For example, trend lines are estimated using OLS. Footnote 21, p. 342, states: "The trend line is (t-statistics in parentheses) . . ."

7

Benston, George J.

"The Validity of Profits-Structure Studies with Particular Reference to the FTC's Line of Business Data

1985(1)

yes

Although the author does not run his own regressions, he reports and provides critique of several empirical studies. Table 3, p. 57, for example, contains coefficient estimates and t-statistics. The author clearly draws conclusions based on these previous studies.

8

McAleer, Michael, Adrian R. Pagan, and Paul A. Volker

"What Will Take the Con Out of Econometrics"

1985(3)

yes

Money demand function regressions with standard errors (Table 3, p. 301). F-statistics (Table 4, p. 303), and specification tests (RESET, normality, heteroskedasticity, etc. (Table 5, p. 304).

9

Schmalensee, Richard

"Do Markets Differ Much?"

1985(3)

yes

"Figure 1 summarizes the results of least squares estimation of equation (1) . . . the number next to each arrow is the probability level at which a standard F-test rejects that restriction" (pp. 345-346).

10

Dewald, William G., Jerry G. Thursby, and Richard G. Anderson

"Replication of Empirical Economics: The Journal of Money, Credit and Banking Project"

1986 (4)

yes

"We obtained estimates qualitatively similar to their estimates in magnitude and statistical significance" (p. 595). T-statistics appear in several tables and coefficients are said to be "significant at the 5 percent level" (Table 3, p. 592).

11

Baumol, William J.

"Productivity Growth, Convergence, and Welfare: What the Long-Run Data Show"

1986(5)

yes

Simple regression of the growth rate of GDP per worker from 1870-1979 on a constant and the log of the level of GDP per worker in 1870 reported along with R-squared (p. 1076).

12

Dominguez, Kathryn M., Ray C. Fair, and Matthew D. Shapiro

"Forcasting the Depression: Harvard versus Yale"

1988(4)

yes

"To study the forecastability of the Depression, we estimate a number of vector autoregression (VAR) models" (p. 599). "It thus seems from at least the two-lag case that the A, B, and C variables are jointly significant" (p. 600).

13

Simon, Carol J.

"The Effects of the 1933 Securities Act on Investor Information and the Performance of New Issues"

1989(3)

yes

Empirical regression results, Table 4 (p. 305), reports variables "statistically different from zero at the 0.10 [and 0.05] level." Tables 6 (p. 306), 7 (p. 309), 8 (p. 311), 9 (p. 312), and 10 (p. 312) also report statistical significance of individual coefficients or F-tests.

14

Blanchard, Olivier Jean, and Danny Quah

"The Dynamic Effects of Aggregate Demand and Supply Disturbances"

1989(4)

yes

VAR standard error bands for variance decompositions reported in Tables 2, 2A, 2B, and 2C (pp. 666-667).

15

Cooley, Thomas F., and Gary D. Hansen

"The Inflation Tax in a Real Business Cycle Model"

1989(4)

yes

Primarily a theoretical paper, but an autoregressive M1 regression with standard errors is reported (p. 741).

Articles omitted from the 1990s survey.

1

Ito, Takatoshi

"Foreign Exchange Rate Expectations: Micro Survey Data"

1990(3)

yes

Coefficients from equation (6) are "estimated by regressing the following eq

2

Stavins, Robert N., and Adam B. Jaffe

"Unintended Impacts of Public Investments on Private Decisions: The Depletion of Forrest Wetlands"

1990(3)

yes

"Equation (24) is a single-equation, fixed-effects model, the parameters of which can be estimated by nonlinear least squares" (p. 344). Table 2 (p. 346) reports coefficient estimates with standard errors.

3

Engel, Charles, and James D. Hamilton

"Long Swings in the Dollar: Are They in the Data and Do Markets Know It?"

1990(4)

yes

Maximum likelihood estimates appear in Table 1 (p. 695) with "Standard errors in parentheses." Table 2 (p. 698) includes "Tests of the null hypothesis that exchange rates follow a martingale against the alternative of segmented trends" with "p-values . . . in parentheses."

4

Bohn, Henning

"Tax Smoothing with Financial Instruments"

1990(5)

yes

VARs estimated and compared to "an ordinary least-squares regression" (p. 1222). Table 1 (p. 1223) notes that "plus signs . . . Indicate rejection of the same null hypothesis at the same, respective, significance levels, based on a regression of tax rates on current returns and four lags of both series."

5

Levin, Sharon G., and Paula E. Stephan

"Research Productivity Over the Life Cycle: Evidence from Academic Scientists"

1991(1)

yes

Tobit regression equations (18) and (19) are on p. 121 while the coefficient estimates with standard errors and statistical significance appear in Table 1 (p. 122) where "a" denotes "statistical significance at 0.10."

6

Ghosh, Atish R., and Paul R. Masson

"Model Uncertainty, Learning, and the Gains from Coordination"

1991(3)

yes

Table 1 (p. 469) reports estimated equations for aggregate demand, money demand, consumer price index, output price change, exchange rate, etc. for two models, along with standard errors for each coefficient.

7

Leitch, Gordon, and J. Ernest Tanner

"Economic Forecast Evaluation: Profit versus the Conventional Error Measures"

1991(3)

yes

Equations (2) and (3) are ARIMA model estimates (p. 584). Tables 2 and 3 (pp. 586-587) report the "absolute value of the t statistic on the slope coefficient."

8

Shiller, Robert J., Maxim Boycko, and Vladimir Korobov

"Popular Attidudes Toward Free Markets: The Soviet Union and the United States Compared"

1991(3)

yes

Survey data, but "Throughout this article, the t statistic is from a probit regression, as described above" (p. 389).

9

Hallman, Jeffrey J., Richard D. Porter, and David H. Small

"Is the Price Level Tied to the M2 Monetary Aggregate in the Long Run?"

1991(4)

yes

Table 1 (p. 847) reports autocorrelations and unit-root tests where "* [denotes] statistically significant at the 95-percent level." Regression results appear throughout the text and Table 2 (p. 851) reports coefficient estimates and standard errors.

10

King, Robert G., Charles I. Plosser, James H. Stock, and Mark W. Watson

"Stochastic Trends and Economic Fluctuations"

1991(4)

yes

Vector error correction model (VECM), along with unit roots tests and unrestricted VAR models. For example, Table 1 (p. 826) includes VAR results as well as p values of unit root tests and estimated cointegrating vectors with p values.

11

Gali, Jordi

"Budget Constraints and Time-Series Evidence on Consumption"

1991(5)

yes

"This approach relies on the assumption that some transformation of labor income follows a univariate ARMA process" (p. 1239). Table 1, p. 1246, reports coefficients, standard errors and p-values.

12

Hercowitz, Zvi, and Michael Sampson

"Output Growth, the Real Wage, and Employment Fluctuations"

1991(5)

yes

VAR models are estimated. Table 2 (p. 1226), for example, reports coefficient estimates with "Standard errors in parentheses."

13

Cecchetti, Stephen G.

"Prices During the Great Depression: Was the Deflation of 1930-1932 Really Unanticipated?"

1992(1)

yes

ARMA models for inflation (and deflation), Dickey-Fuller tests, and a "regression equation" (p. 148-149) for unexpected inflation "estimated by ordinary least squares."

14

Hamilton, James D.

"Was the Deflation During the Great Depression Anticipated? Evidence from the Commodity Futures Market"

1992(1)

yes

Regression and statiscal significance throughout. For example, Table 2 "reports F tests of the null hypothesis [that a parameter is zero] in the regression . . ." (p. 160). Table 3 (p. 161) includes "Regressions of Expected Commodity-Price Inflation" with standard errors and where "* [denotes] Significant at the 5-percent level."

15

Chistiano, Lawrence J., and Martin Eichenbaum

"Current Real-Business-Cycle Theories and Aggregate Labor-Market Fluctuations"

1992(3)

yes

RBC calibration paper that includes both regression and statistical significance. For example, "We cannot reject the indivisible labor model at even the 15-percent significance level" (p. 446). Table 5 (p. 445) contains some estimates that "were set a priori" while "Standard errors (in parentheses) [are reported] only for estimated parameters."

16

Friedman, Benjamin M., and Kenneth N. Kuttner

"Money, Income, Prices, and Interest Rates"

1992(3)

yes

Clearly an empirical paper with regression and statistical significance. Table 1 (p. 474), for example, reports "F statistics" where "Estimated regressions use four lags of each variable" and "* [denotes] Statistically significant a the P<0.05 level."

17

Backus, David K., and Patrick J. Kehoe

"International Evidence on the Historical Properties of Business Cycles"

1992(4)

yes

The data in all of the tables (Tables 1-8, pp. 870-882) are from "The calculations were performed by posing the estimation problem as a generalized-mehtod-of-moments problem using the Hansen-Heaton-Ogaki GAUSS programs. Standard errors are robust to both heteroskedasticity and autocorrelation of the residuals" (p. 870).

18

Wilcox, David W.

"The Construction of U.S. Consumption Data: Some Facts and Their Implications for Empirical Work"

1992(4)

yes

"The variances shown in the second column [of Tables 2, 4, and 5, pp. 930-932] are computed by regressing the log differen of each non-seasonally adjusted sales series on seasonal dummies and differenced day-of-week variables and then squaring the standard error of the regression" (notes to Table 2, p. 930). Integrated moving average models also estimated (p. 932). Statistical significance in text also. For example, ". . . tests of the random-walk hypothesis, as applied to the sampling-error-contaminated data, will reject too often the null hypothesis of no autocorrelation . . ." (p. 934).

19

Burda, Michael C., and Stefan Gerlach

"Intertemporal Prices and the U.S. Trade Balance"

1992(5)

yes

Error correction model is estimated (see p. 1245). Tables 1-4 (pp. 1246-1249) report results where "MSL denotes marginal significance levels. Standard errors are resistant to heteroskedasticity and third-order moving-average errors." In the text, the authors state "While we could not reject at the 5-percent level . . . [that the variables] were nonstationary" (p. 1247).

20

Cooper, Russell, and John Haltiwanger

"The Aggregate Implications of Machine Replacement: Theory and Evidence"

1993(3)

yes

Table 1 (p. 372) reports that "The R-squared values are based on regressions of the reported variable on 12 monthly dummies." Table 2 (p. 377) reports coefficient estimates and "Reported statistics for hypothesis tests are the marginal significance levels from relevant F statistics."

21

Kashyap, Anil K., and David W. Wilcox

"Production and Inventory Control at the General Motors Corporation During the 1920s and 1930s"

1993(3)

yes

Table 1 (p. 397) reports estimates based on an instrumental variables regression with "Standard errors in parentheses; p values are in square brackets." The authors conclude (p. 398) that "the estimated cost of deviating from the target level of inventory is positive in all specifications, though not significantly different from zerio in all sample periods except when we exclude shutdown-contaminated months." Table 2 (p. 399) also reports regression estimates with standard errors and p values.

22

Knetter, Michael M.

"International Comparisons in Pricing-to-Market Behavior"

1993(3)

yes

". . . [T]he regression equations for each destination are estimated jointly, imposing the cross-equation restrictions" (p. 478). Tables 2-8 (pp. 478-482) report coefficient estimats with notes to tables stating that: "Standard errors in parentheses. * Constraint rejected at the 5-percent level."

23

Levine, David I.

"Fairness, Markets, and Ability to Pay: Evidence from Compensation Executives"

1993(5)

yes

Table 2 reports standard errors and t tests for differences in means, while "Regression analysis and nonparametric statistics that address some of the statistical issues ignored in Table 2 are discussed below" (p. 1247).

24

Pesando, James E.

"Art as an Investment: The Market for Modern Prints"

1993(5)

yes

"Econometrically, this procedure is equivalent to regressing the price of each repeat sale on a corresponding time-period dummy. . . " (p. 1077). Table 1 (p. 1078) notes state: "See text for a description of the regression method employed to estimate the log-price index (b). SE(b) is the standard error of the estimated log-price index. . . " Table 2 (p. 1080) notes that "** Statistically significant at the 1-percent level."

25

Feenstra, Robert C.

"New Product Varieties and the Measurement of International Prices"

1994(1)

yes

Estimates for the parameters in equation (10') are "obtained by running weighted least squares (WLS) on (10')." The empirical results are in Table 2 (p. 166) and include coefficient estimates and "standard errors in parentheses." Statistical significance is evident also. For example, "For the two steel products, the constant . . . is not significant" (p. 167).

26

Watson, Mark W.

"Business-Cycle Durations and Postwar Stabilization of the U.S. Economy"

1994(1)

yes

VAR estimated (Table 4, p. 34). T-statistics appear throughout, including Tables 1 and 2 (pp. 28-30), and other tests of statistical significance like the Wilcoxon rank-sum statistic "for comparing the prewar and postwar contraction and expansion phase durations. The statistic is presented in standardized form and can be interpreted like a t statistic for a significant change in the average duration" (p. 26).

27

Gruber, Jonathan

"The Incidence of Mandated Maternity Benefits"

1994(3)

yes

Equation (1), p. 631, shows the "regression equation." Tables 3, 4, 5, and 6 report main results and extensions. Tables contain coefficient estimates and standard errors. Statistical significance is evident also. For example, "it is significant at the 10-percent level" (p. 639).

28

Lewbel, Arthur

"Aggregation and Simple Dynamics"

1994(4)

yes

Tables 1 and 2 (p. 911) contain varying ARIMA specifications with standard errors in parentheses. Significance tests appear throughout. For example, "The coefficient A3 is statistically significant in all the models that include it, while A4 always has a t statistic less than 1. . ." (p. 910)

29

Metrick, Andrew

"A Natural Experiment in 'Jeopardy!'"

1995(1)

yes

Table 7 (p. 249) "contains the results of four logit regressions. . . . Numbers in parentheses are asymptotic t statistics."

30

Bizjak, John M., and Jeffrey L. Coles

"The Effect of Private Antitrust Litigation on the Stock-Market Valuation of Firms"

1995(3)

yes

Tests of statistical significance throughout and regression analysis as well. Table 4 (p. 450) reports "Regressions of Defendant-Firm Abnormal Returns Associated with a Filing." Results include coefficient esimates, t-statistics, and "* Significant at p=0.05."

31

Cogley, Timothy, and James M. Nason

"Output Dynamics in Real-Business-Cycle Models"

1995(3)

yes

Paper contains test statistics for autocorrelation and impulse-response functions (see Table 4, p. 508, for example). Figures contain 95% confidence bands (see Figure 5, p. 504, for example).

32

Noussair, Charles N., Charles R. Plott, and Raymond G. Riezman

"An Experimental Investigation of the Patterns of International Trade"

1995(3)

yes

"The model was estimated for each of the relevant dependent variables, and the results of the estimates are contained in Tables 5, 6, and 7. The standard errors are corrected for heteroskedasticity. . . as well as first-order autocorrelation" (p. 474). "A summary of significance tests of the two models and variables is provided in each of the tables" (p. 476).

33

Cason, Timothy N.

"An Experimental Investigation of the Seller Incentives in EPA's Emission Trading Auction"

1995(4)

yes

Paper uses "two-step generalized least squares estimation" (p. 915). Table 2 (p. 914) reports tests of statistical significance based on this model. Notes to table include "** Statistically significant at the 1-percent level."

34

Lewis, Karen K.

"Occasional Interventions to Target Rates"

1995(4)

yes

Table 2 (p. 700) contains mulinomial logit regressions with coefficient estimates and standard errors in parentheses. The note to the table states that"* Statistically significant at the 5-percent marginal significance level." Several other tables report regression results.

35

Del Boca, Daniela, and Christopher J. Flinn

"Rationalizing Child-Support Decisions"

1995(5)

yes

Table 3 (p. 1255) reports "Ordinary Least-Squares Regression Estimates" with standard errors. For example, "The coefficients associated with both parental incomes are negative; the coefficient associated with the mother's incomeis greater than its standard error . . ." (p. 1255)

36

Bakshi, Gurdip S., and Zhiwu Chen

"The Spirit of Capitalism and Stock-Market Prices"

1996(1)

yes

GMM estimation (Table 3, p. 152) and tests of statistical significance. For example, "The critical t-value, above which the null is rejected, is -1.65 at the 5-percent and -2.33 at the 1-percent significance level (Table 1, p. 149).

37

Ciccone, Antonio, Robert E. Hall

"Productivity and the Density of Economic Activity"

1996(1)

yes

"Table 1 gives the results using the county level education data. The least squares estimate of [gamma] is 1.052 with a standard error of .008" (p. 61). In addition to Table 1 (p. 62), regression results also appear in Table 4 (p. 67), Table 5 (p. 68), and Table 6 (p. 68).

38

Coleman, Wilbur John, II

"Money and Output: A Test of Reverse Causation"

1996(1)

yes

Paper uses a simulated moments estimation technique where the weighting matrix is computed using the Newey-West (1987) procedure. Coefficient estimates are reported in Table 1 (p. 101) although the "standard errors. . . [are] not reported." Statistical significance is reported as well. For example, "The p-value for this statistic is 0.0000, so the estimated model is clearly rejected statistically. . . " (p. 101).

39

Levitt, Steven D.

"How Do Senators Vote? Disentangling the Role of Voter Preferences, Party Affiliation, and Senator Ideology"

1996(3)

yes

Table 3 (p. 433) reports "OLS estimates" (p. 432) that include coefficient estimates and standard errors in parentheses, as well as p-values for year dummies and overidentifying restrictions. Statistical significance is present in text as well. For example, "Instrumenting . . .does in fact reduce the coefficients somewhat in all cases, although the instrumented parameters are not statistically different from the noninstrumented parameters" (p. 434). Tables 5 (p. 436) and 6 (p. 437) also report standard regression results.

40

Lewbel, Arthur

"Aggregation Without Separability: A Generalized Composite Commodity Theorem"

1996(3)

yes

Table 2 (p. 535) reports "Unit Root, Stationarity, and Cointegration Tests" based, in part, on Augmented Dickey-Fuller (1979) tests "of the lagged-level variable in the regression of the differenced variable on a constant, a time trend, and four lags of the differenced variable." Table 3 (p. 537) reports "conintegration tests [which] are Engle-Granger (1987) tests of the null hypothesis that the variables are not cointegrated."

41

Burnside, Craig, and Martin Eichenbaum

"Factor-Hoarding and the Propagation of Business-Cycle Shocks"

1996(5)

yes

"In order to estimate and diagnose the performance of our model we use the generalized method of moments procedure. . . " (p. 1159). The empirical results appear in Table 1 (p. 1162) with coefficient estimates and standard errors. Impulse response functions with standard error bands appear in Figure 4 (p. 1168), Figure 5 (p. 1168), Figure 6 (p. 1170), Figure 7 (p. 1171), and Figure 8 (p. 1172). Statistical significance is also present in the text. For example, "Numbers in the square brackets are probability values associated with Wald statistics for testing the hypothesis that the model and data population values . . . are the same" (p. 1166).

42

Cecchetti, Stephen G., Anil K. Kashyap, and David W. Wilcox

"Interactions Between the Seasonal and Business Cycles in Production and Inventories"

1997(5)

yes

Empirical results appear on pp. 887-890. Table 2 (p. 889) reports "Summary of Empirical Results." Statistical significance appears in the text as well. For example, "The ambiguity arises because we cannot reject either the null of constant seasonal variability over the business cycle of production . . . or the null of no correlation between the inventory interaction and the production seasonal. . . " (p. 890)

43

Gu, Wulong, and Peter Kuhn

"A Theory of Holdouts in Wage Bargaining"

1998(3)

yes

Table 6 (p. 441) reports "Cox Proportional Hazard Coefficients" with "Asymptotic t-statistics in parentheses." Notes to table state "All regressions, except for paper mills only, include 4 dummies for 5 regions. . . " Table 7 (p. 443) reports probit regression results.

44

Banks, James, Richard Blundell, and Sarah Tanner

"Is There a Retirement-Savings Puzzle?"

1998(4)

yes

Regression results, with standard errors, appear in body of paper on pp. 775, 778, 779, and 782. Statistical significance is present as well. For example, "Although there is not a statistically significant difference between retired and unemployed households. . . " (p. 783).

45

Kaminsky, Graciela L., and Carmen M. Reinhart

"The Twin Crisis: The Causes of Banking and Balance-of-Payments Problems"

1999(3)

yes

Graphs including standard error bands are on pp. 481-483. Discussion of Type I and Type II errors on pp. 487-488, and critical values for signaling crisis appear in Table 5 (p. 489).

46

Cawley, John, and Tomas Philipson

"An Empirical Examination in Information Barriers to Trade in Insurance"

1999(4)

yes

Table 3 (p. 838) reports regression coefficient estimates "with t-statistics in parentheses." Statistical significance appears in the text. For example, ". . . But instrumented price was not statistically significant in the AHEAD regressions" (p. 839).

47

Cooper, David J., John H. Kagel, Wei Lo, and Qing Liang Gu

"Gaming Against Managers in Incentive Systems: Experimental Results with Chinese Students and Chinese Managers"

1999(4)

yes

Table 3 (p. 793), Table 4 (p. 797), Table 5 (p. 798) report (probit) regression estimates with standard errors. Statistical significance is readily apparent throughout. For example, "the three context dummies are jointly significant at the 1-percent level" (p. 796), and "the Con3 dummy is positive, relatively large, and statistically significant (p<0.01)" (p. 797).

48

Dahl, Gordon B., and Michael R. Ransom

"Does Where You Stand Depend on Where You Sit? Tithing Donations and Self-Serving Beliefs"

1999(4)

yes

Table 11 (p. 721) and Table 12 (p. 723) report the standard regression output with coefficients, standard errors, and footnotes such as "* Statistically significant at the 5-percent level."

49

Duffy, John, and Jack Ochs

"Emergence of Money as a Medium of Exchange: An Experimental Study"

1999(4)

yes

Experimental data with regression and statistical significance. Table 11 (p. 869), for example, reports "Regression Results for Player Type 1s in Model A" with coefficient estimates and standard errors, as well as footnotes such as "** Significantly different from 0 at the 5-percent level." Tables 12 and 13 (p. 871) report regression results for different player types and models.

50

Evans, William N, Matthew C. Farrelly, and Edward Montgomery

"Do Workplace Smoking Bans Reduce Smoking?"

1999(4)

yes

Table 4 (p. 739), Table 5 (p. 741), and Table 6 (p. 743) report regression results containing coefficient estimates, standard errors, and p-values for "test of overidentifying restrictions" (Table 6, p. 743).

51

Wolfram, Catherine D.

"Measuring Duopoly Power in the British Electricity Spot Market"

1999(4)

yes

Regression and statistical significance are used. For example, "I apply two-state least squares to estimate equation (8)" (p. 820), and "The positive and significant coefficient on QUANTITY is consistent with prior beliefs. . . " (p. 821).

52

Acemoglu, Daron

"Changes in Unemployment and Wage Inequality: An Alternative Theory and Some Evidence"

1999(5)

yes

Although the standard regression output (coefficient estimates and standard errors) does not appear in tables, the paper clearly uses both regression and statistical significance. For example, "consider the regression of real wages on four education dummies, sex, race, a quartic in experience, and a dummy for those living in a metropolitan area. . . " (p. 1273). Also, "In both cases, this drop of over 20 percent is statistically significant at 1 percent" (p. 1272).

53

Davis, Steven J., and John Haltiwanger

"On the Driving Forces Behind Cyclical Movements in Employment and Job Reallocation"

1999(5)

yes

VARs estimated (pp. 1239-1255). Statistical significance as well. For example, Table 2 (p. 1245) reports the serial correlation in the dispersion of industry stock returns, where "*, ** indicates that the correlation differs from zero at the 5-percent (1-percent) significane level in a two-tailed test."

54

Goldberg, Pinelopi Koujianou, and Giovanni Maggi

"Protection for Sale: An Empirical Investigation"

1999(5)

yes

Tables 2-4 (pp. 1148-1150) and Appendix A (p. 1152) report regression results with t-statistics. Statistical significance common. For example, ". . . one might wonder whether we can reject the hypothesis that the government is a pure welfare maximizer (b=1). The answer is yes; even the 99-percent confidence interval does not include b=1" (p. 1147).

55

Lucking-Reiley, David

"Using Field Experiments to Test Equivalence Between Auction Formats: Magic on the Internet"

1999(5)

yes

Regression with standard errors (p. 1075). "The intercept indicates that the revenue difference is not statistically different from zero in experiment ES. . . " (p. 1075).

56

Milyo, Jeffrey, and Joel Waldfogel

"The Effect of Price Advertising on Prices: Evidence in the Wake of 44 Liquormart"

1999(5)

yes

Tables 3-6 (pp. 1092-1094) report regressions, with t-statistics, based on 6,480 observations. Statistical significance is common as well. For example, "Indeed, under each measurement approach, one cannot reject the hypothesis that all effects, save the 'identical product effect,' are identical" (p. 1091).

Omitted articles from the 1980s that use statistical significance but not regression.

1

De Vany, Arthur, and Gail Frey

"Backlogs and the Value of Excess Capacity in the Steel Industry"

1982(3)

no

yes

Tables 2, 3, and 4 contain tests of statistical significance regarding distributions of shipments of steel. For example, ". . . There is insufficient evidence to reject the null hypothesis that net arrivals exhibit a Poisson distribution as assumed in the model " (p. 444).

2

Boskin, Michael J., Marc. S. Robinson, Terrance O'Reilly, and Praveen Kumar

"New Estimates of the Value of Federal Mineral Rights and Land'

1985(5)

no

yes

Confidence levels (Table 4, p. 930), but no regression analysis.

3

Miron, Jeffrey A.

"Financial Panics, the Seasonality of the Nominal Interest Rate, and the Founding of the Fed"

1986(1)

no

yes

No regression results, but statistical significance appears in several places. For example, "The data therefore reject the hypothesis of no change in the frequency of panics at the 99 percent level of confidence" (p. 131).

4

Kroll, Yoram, Haim Levy, and Amnon Rapoport

"Experimental Tests of the Separation Theorem and the Capital Asset Pricing Model"

1988(3)

no

yes

No regression results, but statistical significance appears in several places. For example, "A two-way group by session ANOVA conducted on the slopes of the individual portfolios resulted in a significant difference between the two groups (p<0.05)" (p. 513).

5

Harrison, Glenn W.

"Theory and Misbehavior of First-Price Auctions"

1989(4)

no

yes

Kolmogorov-Smirnov tests (probability) appear for experimental data (Table 2, p. 757).

Omitted articles from the 1990s that use statistical significance but not regression.

1

Cohen, Wesley M., and Steven Klepper

"The Anatomy of Industry R&D Intensity Distributions"

1992(4)

no

yes

No regression results, but statisical significance appears throughout,particularly in Table 2 (p. 787) and Table 3 (p. 790). The authors report that "* A significantly larger number of industries at the 0.05 probability level" (p. 790) and "The signs of all of the correlations are as predicted, with all the estimates significant at the 0.05 level, except . . ." (p. 787).

2

Shogren, Jason F., Seung Y. Shin, Dermot J. Hayes, and James B. Kliebenstein

"Resolving Differences in Willingness To Pay and Willingness To Accept"

1994(1)

no

yes

No regression results, but tables contain many tests of statistical significance, including Table 2, 3, and 4. For example, in Table 2 (p. 262), "* Denotes rejection of H0 at the 5-percent significance level."

3

Hess, Gregory D., and Athanasios Orphanides

"War Politics: An Economic Rational-Voter Framework

1995(4)

no

yes

No regression, but statistical significance using "Fisher's exact test" (p. 836). Table 2 (p. 837), Table 3 (838), Table 4 (p. 839), and Table 5 (p. 840) all report p-values on war frequencies. For example, ". . . The hypothesis of no shift in the unconditional frequency of war is rejected at below the 0.0001 level of statistical significance" (pp. 839-840).

4

Van Boening, Mark V., and Nathaniel T. Wilcox

"Avoidable Cost: Ride a Double Auction Roller Coaster"

1996(3)

no

yes

Table 4 (p. 471) reports z-statistics and p-values for price dispersion across trading periods.

5

Campa, Jose Manuel, and P. H. Kevin Chang

"Arbitrage-Based Tests of Target-Zone Credibility: Evidence from ERM Cross-Rate Options"

1996(4)

no

yes

No regression results, but statistical significance. For example, "Difference of means tests reject at the 1-percent-level the null hypothesis of equal means during these two subperiods for all options and maturities" (p. 731).

Omitted articles from the 1990s that use not regression but not statistical significance.

1

Stockman, Alan C., and Linda Tesar

"Tastes and Technology in a Two-Country Model of the Business Cycle: Explaining International Comovements"

1995(1)

yes

no

Mostly a calibration exercise, but regression is used to estimate some of the model parameters. For example, to find the elasticity between traded and nontraded goods, "we regress the nontraded-good expenditure share on the price index for nontraded goods and include per capita GDP to pick up income effects" (p. 178).