Ralf Steinhauser

Research

Published Papers

"A Framework for Forecasting the Path of US CO2 Emissions Using State-Level Information", with Maximilian Auffhammer,Review of Economics and Statistics, forthcoming.

Abstract: In this paper we compare the most common reduced form models used for emissions forecasting, point out shortcomings and suggest improvements. Using a U.S. state level panel data set of CO2 emissions we test the performance of existing models against a large universe of potential reduced form models. Our preferred measure of model performance is the squared out-of-sample prediction error of aggregate CO2 emissions. We find that leading models in the literature, as well as models selected based on an emissions per capita loss measure or different in-sample selection criteria, perform significantly worse compared to the best model chosen based directly on the out-of-sample loss measure defined over aggregate emissions. Unlike the existing literature, the tests of model superiority employed here account for model search or `data snooping' involved in identifying a preferred model. Forecasts from our best performing model for the United States are 100 million tons of carbon lower than existing scenarios predict.

"Managers' Behavior under Weak Governance.", Forthcoming in Abol Jalilvand and Tassos Malliaris (eds.)Risk Management and Corporate Governance, (Routledge), 2011.

Abstract:It is conventional wisdom that managers who are not closely monitored pursue personal goals rather than maximize shareholder wealth. Yet little is known about what these goals are, despite the importance of understanding manager behavior when designing corporate governance rules. I provide new insights into managers’ personal preferences by studying the variations in corporate environmental and social performance associated with different corporate governance provisions. I employ a unique dataset on corporate governance and corporate social responsibility and exploit variations in takeover defenses to analyze differences in managers’ behavior. I find that with weaker governance, more resources are allocated into environmentally and socially responsible objectives and away from core responsibilities. These findings support a theory that managers want to live a “good life”.

"The Future Trajectory of US CO2 Emissions: The Role of State vs. Aggregate Information.", with Maximilian Auffhammer,Journal of Regional Science2007,
47(1), pp. 47-61.

Abstract:This paper provides comparisons of a variety of time series methods for short run forecasts of the main greenhouse gas, carbon dioxide, for the United States, using a recently released state level data set from 1960-2001. We test the out-of-sample performance of univariate and multivariate forecasting models by aggregating state level forecasts and forecasting the aggregate directly. We find evidence that forecasting the disaggregate series and accounting for spatial effects drastically improves forecasting performance under Mean Squared Error Loss. Based on the in-sample observations we attempt to explain the emergence of voluntary efforts by states to reduce greenhouse gas emissions. We find evidence that states with decreasing per capita emissions and a "greener" median voter are more likely to push towards voluntary cutbacks in emissions.

"Tax Smoothing and the Cross-Country Pattern of Privatization.", with Peter Berck and Jonathan Lipow.World Development2006; 34(2) pp. 238-246. .

Abstract: Privatization generally enhances firm-level efficiency. The impact of privatization on fiscal efficiency has, however, been overlooked. Using the ‘‘tax-smoothing’’ ideas articulated by R. Barro (1979) [On the determination of the public debt. Journal of Political Economy, 87, 940–971] and by H. Bohn (1990) [Tax smoothing with financial instruments. American Economic Review, 80, 1217–1230], we argue that privatization may have an important impact on the welfare losses associated with tax collection. This impact can either enhance or erode the efficiency of taxation. We hypothesize that countries that benefit in terms of fiscal efficiency will privatize aggressively while countries that enjoy fiscal benefits as a result of state ownership will show little interest in privatization. Statistical tests are conducted that provide considerable support for the hypothesis.

Working Papers

"Does anybody give a dam? The importance of public awareness for urban water conservation during drought"(with Emma Aisbett) ,  March 2011   

Abstract: Demand management has been of interest in dry climates such as Australia, Spain and the Western United States for decades. It is particularly important to understand policy options during drought conditions, as drought periods have a disproportionate effect on supply infrastructure decisions. While water-conservation campaigns aimed at inducing voluntary consumption reductions are almost universally employed by water managers in times of supply constraint, voluntary measures are generally dismissed in the economics literature as ineffective. We argue that the robust positive correlation between dam levels and consumption after controlling for policy changes suggests that there is a significant component of voluntary conservation. Furthermore, omitting dam levels from regressions may bias estimated impacts of policy changes.

Keywords: water use, demand management, pricing, behavioral aspects

"Principals vs. Principles: What Do Managers Do When Governance Is Slack?", October 2009

Abstract: The separation of ownership and control in corporations opens up the potential for moral hazard. When designing corporate governance rules it is essential to understand how these diverging goals may affect managers’ behavior. This paper looks at managers’ preferences for corporate social responsibility (CSR) firm outcomes. We argue that managers will derive utility from `doing good’ using firms resources. We provide a theoretical framework to show managers and shareholder tradeoffs in their preferences for the firm’s optimal level of CSR and derive testable hypothesises on expected changes in CSR performance based on changes in corporate governance. We empirically test these using a firm level dataset on CSR, corporate governance and Compustat firm indicators. We find evidence that entrenched managers divert resources away from core firm responsibilities toward CSR areas which give managers more utility from `doing good’.  

Keywords: Corporate Social Responsibility, Corporate Governance, Antitakeover Amendments, Firm, Performance JEL Codes: M14, G34, Q20

"Bayesian Model Averaging vs. Out-of-sample Model Selection: An Application to Forecasting Carbon Emissions", May 2005

Abstract: In this paper we compare forecast for U.S. CO2 emissions from the frequentist out-of-sample approach based on the Reality Check by White (2000) and Bayesian model averaging (BMA) with each other, and with benchmark models in the literature. Both methods perform well in the short term. Compared to the out-of-sample approach BMA has a tendency to put significant weight on models with higher order polynomial terms, a feature that leads to unrealistic longer-term forecasts. This finding contributes to the ongoing discussion in the econometric literature over the benefits of in-sample verses out-of-sample model selection and testing. We end by suggesting an out-of-sample adaptation of the BMA approach which has potential advantages over either of the existing methods.

Keywords: Forecasting, Bayesian Model Averaging, Out-of-sample Model Selection, Climate Change, United States, CO2 Emissions, Data Snooping

"Efficiency and Information in Markets - A Laboratory Experiment on the Effect of Insider Information in Markets", (in German) with Ben Greiner and Charlotte Möser, August 2000

Abstract: The strong form of market efficiency cannot be studied in real markets where you don’t observe insider information. We simulate a stock market laboratory with and without insider traders and compare the information efficiency in both markets. We look at the change of behavior of traders in asset markets when it is known that there is private information in the market. Our results support the semi strong market efficiency hypothesis. We further investigate how insider information is valued. A sealed bid second price auction is used to allocate information of next period’s news announcements. We find that insiders make above average profits but also often fall for the winner’s course.

Work in Progress

"Overconfidence under pressure: Self-assessment with real incentives"

Abstract: Two studies investigate students’ overconfidence about their own performance in class. The first study asks students - in an introductory quiz - questions with varying degree of difficulty before asking them to assess their expected relative outcome in the class. We find no influence of more challenging questions on their otherwise self-confident assessment of their relative performance. The second study looks at students’ mid-semester expectations of their end-of-semester relative performance. In one setting students are primed by a lecture on overconfidence before they are asked to judge their own expected relative outcome at the end of the course. The same group of students is asked later in an exam setting the same question again but now with a reward of extra class credit offered if they guess close enough to their final standing in class. Our results suggest that when students are faced with a real incentive for a best guess of their final relative standing they overcome some of their overconfidence and guess more accurately. Nevertheless signs of overconfidence still dominate students’ answers with 57% guessing a higher outcome than they achieve. The average individual student guess is 28% above the average students’ outcome and optimistic vs. pessimistic beliefs of 5.5:1. Even with the non-monetary incentives offered students still exhibit strong overconfident beliefs.

Keywords: Overconfidence, Overly Optimistic Beliefs, Self-confidence, Priming

"Does raising teacher pay boost student achievement? A regression discontinuity approach"(with Andrew Leigh and Damien Moore)

A central question in the economics of education is the impact of teacher salary on student achievement. Previous studies have often struggled to deal with problems of endogeneity – administrators and voters may punish teachers for poor student performance by keeping teacher salaries low. Using data from the state of Western Australia, we exploit an unusual feature of the system: salaries are uniform across the state, except that schools outside the capital city pay higher salaries to their teachers. The salaries rise in levels across districts whose boundaries are drawn along longitudes and latitude.  Our approach assumes that schools which are very close to these geographic boundaries, but on opposite sides, should be identical in all respects except for the salary differential. This allows us to employ a regression discontinuity approach to identify whether literacy and numeracy scores increase with teacher salaries.

Keywords: educational production function, regression discontinuity, school expenditure, teacher labor markets JEL Codes: I21, I22

 “Social preferences in the real world: A stated preference analysis" (with Thomas Shael)

Abstract: This paper employs an innovative survey technique to analyse social preferences. We discuss the advantages of a stated preference methodology over laboratory experiments studying preferences. A survey of 650 ANU undergraduate students is conducted and used to test which of the prominent social preference models from the literature explains the stated preference observations best. Our results confirm the neoclassical predictions can’t account for people’s preferences. People behave to a large extend altruistically in our experiment and we find evidence for fairness considerations. Charness and Rabin’s (2002) social welfare preference model explains the data best. Our survey technique allows us to look at how people’s altruism changes when the outcomes are reduced cost vs. increased benefits. We find that individuals are more selfish when their outcomes are in the loss domain compared to when they are in the gain domain. Similarly subjects are more selfish when the outcomes of others that they trade off against are in the gain domain versus the loss domain. These findings are explained with reference-dependent preferences.

 Keywords: Social Preferences, Reference –dependent Preferences, Stated Preference, Difference Aversion, Fairness, Inequity Aversion  JEL Codes: A13, C91, C93, D63

"Does Financial Literacy Boost Savings? An Instrumental Variables Approach" (with Emma Aisbett)

"Estimating the Impact of Fiscal Stimulus on Household Expenditure" (Markus Brueckner, Liyi Pan and Emma Aisbett)