UK manufacturing productivity has declined sharply since 2008, yet little is known about the role of trends in firm-level markups in this context. We estimate firm-level markups using a structural demand approach with revenue shares and industry elasticities, which is complemented by a primal method. Findings indicate a decline in aggregate markups of two to five per cent from 2008 to 2019, driven mainly by within-firm declines. These results suggest that the UK's productivity slowdown reflects structural deterioration within manufacturing firms rather than reallocation effects.

Difference-in-Differences and Event-Study regression estimators are commonly used to estimate treatment effects. However, when effects are heterogeneous, the standard two-way fixed effects (TWFE) regression can provide biased estimates. Recent literature proposes alternative estimators to overcome the so-called “bad comparisons” problem, allowing unbiased estimates of the treatment effect to be recovered. To date, attention has primarily focused on linear models despite the prevalence of other outcome types in applied research. In this paper, we address this gap by extending five of these alternative estimators for estimating treatment effects on count and binary outcomes and exploring their relative performance within a simulation exercise where true effects are known. Our simulation results indicate that while these estimators can produce unbiased estimates for linear outcomes, they may fail to recover the true treatment effect for nonlinear outcomes if used ‘straight out of the box’. We show that Interaction-weighted, IPW, and Extended-TWFE estimators can be readily adjusted to provide unbiased estimates. Finally, we apply these estimators to prior published work examining the effect of star coauthorship on peers’ productivity using citation data, revealing moderate differences in the magnitude of the effects across the estimators.

We build on Modestino et al., who analysed "opportunistic upskilling" during the 2007 recession in the US, by extending their approach to European labour markets following recent economic shocks. Using a comprehensive dataset of online job postings (2019-2023) and an instrumental variable approach leveraging Ukrainian refugee inflows, we find that increases in labour supply have a significant positive effect on both education and experience requirements. Our findings reinforce the idea that employers opportunistically upskill labour demand in different but challenging macroeconomic contexts. The identification of such patterns, however, depends on the presence of a plausibly exogenous shock to labour markets.

There is increasing interest in recruiting star scientists to catalyze clusters in targeted research areas. While star-arrival effects on incumbents' productivity are often positive, we know less about how the relatedness between incumbents' and stars' knowledge intermediates the size of these effects. Using Scopus data on publications and citations, we find that incumbent productivity is insensitive to a fixed measure of relatedness at the time of arrival but is positively affected when relatedness changes after arrival. This distinction between "being related" and "becoming related" to the star matters. Results are robust across alternative measures of relatedness, output, and time windows.

The task-based approach has become the dominant framework for studying the labor-market effects of artificial intelligence (AI), typically emphasizing the replacement of human workers by machines. Motivated by growing empirical evidence that contemporary AI is more often used as a tool that augments workers, this paper develops two related task-based models in which AI enhances worker productivity without automating tasks. Abstracting from capital, we develop a pair of related task-based models that examine how technological progress in AI that provides new tools to augment workers affects aggregate productivity and wage inequality. Both models emphasize the role of human capital in intermediating the effects of AI-related technological shocks. In the first model, AI use requires specialized expertise, and technological progress expands the set of tasks for which such expertise is effective. We show that a larger supply of AI expertise amplifies the productivity gains from improvements in AI technology while attenuating its adverse effects on wage inequality. The second model focuses on non-AI skills, allowing AI tools to alter the set of tasks that workers can perform given their skills. In equilibrium, workers allocate across tasks in response to wages, generating an endogenous distribution of skills across the task space. A central result is that aggregate productivity and wage inequality depend on different global properties of this equilibrium distribution: productivity is particularly sensitive to thinly staffed tasks that create bottlenecks, while wage inequality is driven by the concentration of workers in a narrow set of tasks. As a result, improvements in AI tools can induce non-monotonic co-movement between productivity and inequality. By linking these mechanisms to multidimensional human capital---including AI expertise and higher-order non-AI skills---the paper highlights the role of education and training policies in shaping the economic consequences of AI-driven technological change. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

We explore the impact of artificial intelligence (AI) on the knowledge production function. We characterize AI as a tool, not for full automation but rather for augmentation through enhanced search over combinatorial spaces. This leads to increased scientific productivity. We decompose knowledge production into a multi-stage process to shed light on the "jagged frontier" of AI in science, revealing differential returns to different tools across domains (e.g., data-rich biology vs. anomaly-sparse physics) and workflow stages (e.g., strong design aids like AlphaFold vs. subtler question generation tools). We treat human judgment as indispensable for tasks involving abductive inference, contextual nuance, and trade-offs, particularly in data-sparse environments. Drawing on a task-based model that distinguishes "ordinary" from AI-expert scientists, we describe how exogenous improvements in AI yield nonlinear productivity gains amplified by the share of scientists that are AI-experts to underscore the role of AI complements like skills training and organizational design. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

This paper contributes to the growing literature on the impact of connections to star scientists on the productivity of academic scientists. The existing literature generally focuses on larger economies and specific scientific fields in evaluating star-connection effects. It has rarely examined the particular channels through which stars have their effects. Using natural language processing (NLP) techniques to explore the acknowledgement texts of a broad corpus of published papers from three small open economies, we examine the effects of star help revealed by the acknowledgement texts published in articles. Using an event-study framework with matched data, we find evidence of an economically and statistically significant effect on scientist productivity in the year of acknowledgement of star help. However, there is only evidence of an enduring productivity effect if scientists maintain their acknowledgement of ties to the star over time. A similar pattern is evident across different types of acknowledgements, except for acknowledgements of star help with access to materials, which shows an enduring effect even after a single acknowledgement. The largest estimated star-help effects are found for authors in lower quartiles of the field-specific productivity distribution measured in the year before the help is acknowledged. The results are robust to using a raw-publications-based measure of scientist productivity in place of our preferred citation-weighted publications measure of productivity, suggesting that the observed productivity effect is unlikely to be due to a pure signalling effect. We discuss the implications of these findings for the design of star recruitment and integration policies.

•The “greened” savings variants (Green and GS) offered only limited support for weaker theoretical propositions.•Increasingly comprehensive savings variants were good predictors of future consumption changes.•The β1 coefficients of technology-augmented savings ranged from 0.79 to 1.05, with economic theory predicting a value of 1•The ComprehensiveTFP variant, generally held the smallest standard errors and the tightest β1 coefficient range.•Even the strongest theoretical proposition of a one-for-one relationship between savings and future consumption could not be rejected over some time horizons.

We model a key step in the innovation process, hypothesis generation, as the making of predictions over a vast combinatorial space. Traditionally, scientists and innovators use theory or intuition to guide their search. Increasingly, however, they use artificial intelligence (AI) instead. We model innovation as resulting from sequential search over a combinatorial design space, where the prioritization of costly tests is achieved using a predictive model. The predictive model's ranked output is represented as a hazard function. Discrete survival analysis is used to obtain the main innovation outcomes of interest – the probability of innovation, expected search duration, and expected profit. We describe conditions under which shifting from the traditional method of hypothesis generation, using theory or intuition, to instead using AI that generates higher fidelity predictions, results in a higher likelihood of successful innovation, shorter search durations, and higher expected profits. We then explore the complementarity between hypothesis generation and hypothesis testing; potential gains from AI may not be realized without significant investment in testing capacity. We discuss the policy implications.

Understanding the poor productivity performance of the UK economy since the financial crisis is complicated by the well-known challenges in estimating total factor productivity (TFP) using only revenue data. We develop a structural framework to infer quality-adjusted TFP from an estimated firm-level revenue function. We use microdata for two sectors previously identified as being significant contributors to the UK's productivity growth slowdown-manufacturing and ICT-from 2008 to 2019. The revenue function is estimated using the Blundell-Bond System GMM estimator. We also use an alternative cost-shares approach to identifying and measuring TFP. For both methods, we find an overall fall in TFP levels in manufacturing and a rise in ICT. We find a striking decline of between 13% and 18% in the level of within-firm manufacturing TFP, and of between 11% and 16% in ICT, although with reallocation effects differing between the two sectors. The finding of declining within-firm TFP is robust, although the magnitude varies between methods. We discuss a possible explanation for this extended UK productivity puzzle based on the relative underperformance of UK firms in international markets.

There is increasing interest in star recruitment policies as a means to catalyze the development of clusters in targeted research areas. While there is growing evidence of positive star-arrival effects, we know less about how relatedness to the arriving star intermediates the size of the productivity effect on co-located peers. In the context of scientific productivity, we use various relatedness indexes to estimate this “intensity of treatment effect,” where the intensity is proxied by the relatedness measure. Recognizing that greater relatedness can both increase absorptive capacity and exacerbate potential knowledge redundancy, we allow for a possible non-linearity between the relatedness measure and the observed productivity effect on incumbent scientists. We find evidence of an inverted U-shaped relationship. Interestingly, an asymmetric relatedness measure calculated from the perspective of the star provides a better fit to the data than one calculated from the perspective of the incumbent scientist, which we interpret as indicating the importance of the star’s engagement with their new colleagues. Our results are robust to using various relatedness measures based on journal and keyword overlap between the treated scientist and the star. We discuss the possible implications for the design of star recruitment policies.

There is increasing policy interest in the recruitment and integration of star scientists as a mechanism to catalyse research productivity. We use rich data for three Small Open Economies (Ireland, Denmark, and New Zealand) on publications, citations and co-authorships to examine how co-authorship with a co-located star scientist affects the co-author's productivity, both including and excluding the output directly co-authored with the star. The latter effect provides a measure of the extent to which star collaborations crowd out/in other output. Event-study analyses reveal that star co-authorships are associated with economically and statistically significant increases in co-authors' output (measured by field-normalized total citations). Output in the three years after the initial star co-authorship is increased by 89.6 % when star co-authored publications are included and by 16.2 % when they are excluded. The results are robust to using an alternative measure of quality-adjusted output based on journal publication quality. We find co-authoring with a star increases the quality but not quantity of output when star co-authored publications are excluded. We explore heterogeneity by period, field and whether the authors have multiple star co-authorships. We conclude that policymakers' and institutions' efforts to promote access to star scientists may have substantial direct and indirect effects on the productivity of incumbent scientists within departments.

New product innovation in fields like drug discovery and material science can be characterized as combinatorial search over a vast range of possibilities. Modeling innovation as a costly multi-stage search process, we explore how improvements in artificial intelligence (AI) could affect the productivity of the discovery pipeline in allowing improved prioritization of innovations that flow through that pipeline. We show how AI-aided prediction can increase the expected value of innovation and can increase or decrease the demand for downstream testing, depending on the type of innovation, and examine how AI can reduce costs associated with well-defined bottlenecks in the discovery pipeline.

There is increasing interest among policymakers in small open economies in the use of star-scientist recruitment policies to catalyse the development of local clusters in targeted research areas. We use Scopus to assemble a dataset on over 1.4 million publications and subsequent citations for Denmark, Ireland and New Zealand from 1990 to 2017. An event-study model is used to estimate the dynamic effects of a star arrival on quality-adjusted research output at both the department and matched individual incumbent levels. Star arrivals are associated with statistically significant increases in department output (excluding the output of the star) of between 12% and 25% after 4 years. At the incumbent level, star arrivals lead to an approximately 5% increase in individual output, with substantially larger increases for incumbents who co-author with the star.

This paper constructs measures of Genuine Savings (GS), the leading economic indicator of sustainable development, for 28 European economies from 1990 to 2016. The World Bank publishes GS estimates for most countries in the world. The World Bank's estimates are extended to include local air pollutants (sulfur dioxide, non-metallic volatile organic compounds, nitrogen oxides and ammonia), on a country-specific basis. The extended pollution damages have a sizeable impact, even for highly developed economies. As many as 11 member states signaled persistent unsustainable development compared with just two in the equivalent World Bank dataset. Overall, it appears economies with higher levels of per capita national income and those more open to trade tend to hold higher savings rates. Findings of negative savings, an indicator of unsustainable development, were concentrated in the former communist regimes of Eastern Europe.

After a century of Irish independence, this study constructs long run Genuine Savings estimates, a leading economic indicator of sustainable development, to reassess Irish economic history from the vantage of sustainable development. The main difference uncovered surrounds the post-1950 period where Ireland failed to achieve economic convergence and was considered an economic failure in growth terms. From a sustainability perspective, Ireland may have been an overachiever during a "great transition" of sustainable development driven by improved institutions and policies. The findings show the value of the sustainable development perspective in shedding new light on a country's development experience.

This paper relies on a microsimulation framework to undertake an analysis of the distributional implications of the COVID-19 crisis over three waves. Given the lack of real-time survey data during the fast moving crisis, it applies a nowcasting methodology and real-time aggregate administrative data to calibrate an income survey and to simulate changes in the tax benefit system that attempted to mitigate the impacts of the crisis. Our analysis shows how crisis-induced income-support policy innovations combined with existing progressive elements of the tax-benefit system were effective in avoiding an increase in income inequality at all stages of waves 1-3 of the COVID-19 emergency in Ireland. There was, however, a decline in generosity over time as benefits became more targeted. On a methodological level, our paper makes a specific contribution in relation to the choice of welfare measure in assessing the impact of the COVID-19 crisis on inequality.

This chapter examines national science policy as a case-study in evidence-based policy design. Its reviews the strategy and science of Irish science policy in light of the challenges for such policies in an SOE. The success of knowledge intensive industries depends on access to knowledge. However, private firms tend to underinvest in basic science where much of the benefit spills over to other firms, highlighting an important role for governments. Governments of SOEs face two challenges in devising a strategy for science policy: first, the benefits of science investments are likely to flow disproportionately to other countries; second, small size may limit the benefits of agglomeration economies that are central to many knowledge-intensive industries. Despite obvious spillover and scale challenges – geographical stickiness of new knowledge production and the capacity to absorb knowledge from the global stock depends on being active at the frontiers of knowledge production. The chapter concludes that the national benefit of research is the advantage in being able to access knowledge produced elsewhere.