UK Macroeconomic Forecasting with Many Predictors: Which Models Forecast Best and When Do They Do So?

Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting models as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output growth and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.

Optimal contracting with private information on cost expectation and variability

We study the screening problem that arises in a framework where, initially, the agent is privately informed about both the expected production cost and the cost variability and, at a later stage, he learns privately the cost realization. The speci c set of relevant incentive constraints, and so the characteristics of the optimal mechanism, depend nely upon the curvature of the principal s marginal surplus function as well as the relative importance of the two initial information problems. Pooling of production levels is optimally induced with respect to the cost variability when the principal's knowledge imperfection about the latter is sufficiently less important than that about the expected cost.

UK Macroeconomic Forecasting with Many Predictors: Which Models Forecast Best and When Do They Do So?

Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting model as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.

On the time-varying relationship between EMU sovereign spreads and their determinants

We use a dynamic multipath general-to-specific algorithm to capture structural instability in the link between euro area sovereign bond yield spreads against Germany and their underlying determinants over the period January 1999 – August 2011. We offer new evidence suggesting a significant heterogeneity across countries, both in terms of the risk factors determining spreads over time as well as in terms of the magnitude of their impact on spreads. Our findings suggest that the relationship between euro area sovereign risk and the underlying fundamentals is strongly timevarying, turning from inactive to active since the onset of the global financial crisis and further intensifying during the sovereign debt crisis. As a general rule, the set of financial and macro spreads’ determinants in the euro area is rather unstable but generally becomes richer and stronger in significance as the crisis evolves.

My Group Beats Your Group: Evaluating Non-Income Inequalities

This paper proposes a new methodology, the Domination Index, to evaluate non-income inequalities between social groups such as inequalities of educational attainment, occupational status, health or subjective well-being. The Domination Index does not require specific cardinalisation assumptions, but only uses the ordinal structure of these non-income variables. We approach from an axiomatic perspective and show that a set of desirable properties for a group inequality measure when the variable of interest is ordinal, characterizes the Domination Index up to a positive scalar transformation. Moreover we make use of the Domination Index to explore the relation between inequality and segregation and show how these two concepts are related theoretically.

Genuine savings and future well-being in Germany, 1850-2000

Genuine Savings (GS), also known as ‘net adjusted savings’, is a composite indicator of the sustainability of economic development. Genuine Savings reflects year-on-year changes in the total wealth or capital of a country, including net investment in produced capita, investment in human capital, depletion of natural resources, and damage caused by pollution. A negative Genuine Savings rate suggests that the stock of national wealth is declining and that future utility must be less than current utility, indicating that economic development is non-sustainable (Hamilton and Clemens, 1999). We make use of data over a 150 year period to examine the relationship between Genuine Savings and a number of indicators of well-being over time, and compare the relative changes in human, produced, and components of natural capital over the period. Overall, we find that the magnitude of genuine savings is positively related to changes in future consumption, with some evidence of a cointegrating relationship. However, the relationships between genuine savings and infant mortality or average heights are less clear.

Can Early Intervention Policies Improve Well-being? Evidence from a randomized controlled trial

Many authors have proposed incorporating measures of well-being into evaluations of public policy. Yet few evaluations use experimental design or examine multiple aspects of well-being, thus the causal impact of public policies on well-being is largely unknown. In this paper we examine the effect of an intensive early intervention program on maternal well-being in a targeted disadvantaged community. Using a randomized controlled trial design we estimate and compare treatment effects on global well-being using measures of life satisfaction, experienced well-being using both the Day Reconstruction Method (DRM) and a measure of mood yesterday, and also a standardized measure of parenting stress. The intervention has no significant impact on negative measures of well-being, such as experienced negative affect as measured by the DRM and global measures of well-being such as life satisfaction or a global measure of parenting stress. Significant treatment effects are observed on experienced measures of positive affect using the DRM, and a measure of mood yesterday. The DRM treatment effects are primarily concentrated during times spent without the target child which may reflect the increased effort and burden associated with additional parental investment. Our findings suggest that a maternal-focused intervention may produce meaningful improvements in experienced well-being. Incorporating measures of experienced affect may thus alter cost-benefit calculations for public policies.

Decision Errors, Legal Uncertainty and Welfare: a General Treatment

This paper provides a general treatment of the implications for welfare of legal uncertainty. We distinguish legal uncertainty from decision errors: though the former can be influenced by the latter, the latter are neither necessary nor sufficient for the existence of legal uncertainty. We show that an increase in decision errors will always reduce welfare. However, for any given level of decision errors, information structures involving more legal uncertainty can improve welfare. This holds always, even when there is complete legal uncertainty, when sanctions on socially harmful actions are set at their optimal level. This transforms radically one’s perception about the “costs” of legal uncertainty. We also provide general proofs for two results, previously established under restrictive assumptions. The first is that Effects-Based enforcement procedures may welfare dominate Per Se (or object-based) procedures and will always do so when sanctions are optimally set. The second is that optimal sanctions may well be higher under enforcement procedures involving more legal uncertainty.