Knowledge that Transforms

We verify whether parameter-driven and observation-driven classes of dynamic models can outperform each other in predicting time-varying parameters. We consider existing and new dynamic models for counts and durations, but also for volatility, intensity, and dependence parameters. In an extended Monte Carlo study, we present evidence that observation-driven models based on the score of the predictive likelihood function have similar predictive accuracy compared to their correctly specified parameter-driven counterparts. Dynamic observation-driven models based on predictive score updating outperform models based on conditional moments updating. Our main findings are supported by the results from an extensive empirical study in volatility forecasting.

We estimate a Stock/Watson index of economic activity to assess U.S. business cycle volatility since 1867. We replicate the Great Moderation of the 1980s and 1990s and find exceptionally low volatility also in the Golden Age of the 1960s. Postwar moderation relative to pre-1914 occurs under constant but not time-varying factor loadings, suggesting structural change toward more volatile sectors. For comparable series, the U.S. postwar business cycle was as volatile overall as under the Classical Gold Standard, but much less so during the Great Moderation and the Golden Age.

We consider the link between poverty and subjective wellbeing and focus in particular on potential adaptation to poverty. We use panel data on almost 54,000 individuals living in Germany from 1985 to 2012 to show, first, that life satisfaction falls with both the incidence and intensity of contemporaneous poverty. We then reveal that there is little evidence of adaptation within a poverty spell: poverty starts bad and stays bad in terms of subjective well-being. We cannot identify any cause of poverty entry that explains the overall lack of poverty adaptation.

We find that more closely related populations are more prone to engage in international conflict with each other. We provide an economic interpretation based on two connected mechanisms. First, more closely related groups share more similar preferences over rival goods and are thus more likely to fight over them. Second, rulers have stronger incentives to conquer populations more similar to their own, to minimize postconflict heterogeneity in preferences over government types and policies. We find support for these mechanisms using evidence on international conflicts over natural endowments and on territorial changes, including decolonization.

Political regimes influence the content of education and criteria used to select and evaluate students.We study the impact of a socialist education on the likelihood of obtaining a college degree and on several labor market outcomes by exploiting the reorganization of the school system in East Germany after reunification. Our identification strategy uses cutoff birth dates for school enrollment that lead to variation in the length of exposure to the socialist education system within the same birth cohort. An additional year of socialist education decreases the probability of obtaining a college degree and affects longer-term labor market outcomes for men.

We compare the twelve-month default probability among subprime borrowers differing only in the number of months before their first lump-sum property tax payment, after which time they may be exposed to reduced liquidity. We show that borrowers with an earlier property tax bill—within three months of origination—have 2% to 6% higher first-year default rates than borrowers facing their first property tax bill ten to twelve months after origination. Lump-sum property tax payments appear to produce a persistent state of low liquidity, the length of which raises the likelihood of default. These results are about one-third the effect size of a transition from 10% positive to 20% negative equity found in the literature. This paper provides causal evidence that liquidity constraints are important predictors of mortgage default.

We consider constructing probability forecasts from a parametric binary choice model under a large family of loss functions (“scoring rules”). Scoring rules are weighted averages over the utilities that heterogeneous decision makers derive from a publicly announced forecast (Schervish, 1989). Using analytical and numerical examples, we illustrate howdifferent scoring rules yield asymptotically identical results if the model is correctly specified. Under misspecification, the choice of scoring rule may be inconsequential under restrictive symmetry conditions on the data-generating process. If these conditions are violated, typically the choice of a scoring rule favors some decision makers over others.

We study the role of labor inputs in religious attendance using data on Oklahoma Methodist congregations from 1961 to 2003. Pastors play a significant role in church growth: replacing a 25th percentile pastor with a 75th percentile one increases annual attendance growth by 3%. A pastor’s performance in his or her first church (largely the result of random assignment) predicts future performance, suggesting a causal effect of pastors on growth. The deployment of pastors by the church indicates efficient use of labor: low-performing pastors are more likely to be rotated or exit the sample, and high-performing pastors are moved to larger congregations.

In the 1980s and 1990s, the U.S. labor market experienced a remarkable polarization along with fast technological catch-up as Europe and Japan improved their global innovation performance. Is foreign technological convergence an important source of wage polarization? To answer this question, we build a multicountry Schumpeterian growth model with heterogeneous workers, endogenous skill formation, and occupational choice. We show that convergence produces polarization through business stealing and increasing competition in global innovation races. Quantitative analysis shows that these channels can be important sources of U.S. polarization. Moreover, the model delivers predictions on the U.S. wealth-income ratio consistent with empirical evidence.

This paper examines how risk behavior in the limelight differs from that in anonymity. In two separate experiments, we find that subjects are more risk averse in the limelight. However, risky choices are similarly path dependent in the different treatments. Under both limelight and anonymous laboratory conditions, a simple prospect theory model with a path-dependent reference point provides a better explanation for subjects’ behavior than a flexible specification of expected utility theory. In addition, our findings suggest that ambiguity aversion depends on being in the limelight, that passive experience has little effect on risk taking, and that reference points are determined by imperfectly updated expectations.