Hasil untuk "Labor policy. Labor and the state"

David McDowall, R. McCleary, E. Meidinger et al.

Matthias Mertens, Adam Kuzee, Brittany S. Harris et al.

We propose that AI automation is a continuum between: (i) crashing waves where AI capabilities surge abruptly over small sets of tasks, and (ii) rising tides where the increase in AI capabilities is more continuous and broad-based. We test for these effects in preliminary evidence from an ongoing evaluation of AI capabilities across over 3,000 broad-based tasks derived from the U.S. Department of Labor O*NET categorization that are text-based and thus LLM-addressable. Based on more than 17,000 evaluations by workers from these jobs, we find little evidence of crashing waves (in contrast to recent work by METR), but substantial evidence that rising tides are the primary form of AI automation. AI performance is high and improving rapidly across a wide range of tasks. We estimate that, in 2024-Q2, AI models successfully complete tasks that take humans approximately 3-4 hours with about a 50% success rate, increasing to about 65% by 2025-Q3. If recent trends in AI capability growth persist, this pace of AI improvement implies that LLMs will be able to complete most text-related tasks with success rates of, on average, 80%-95% by 2029 at a minimally sufficient quality level. Achieving near-perfect success rates at this quality level or comparable success rates at superior quality would require several additional years. These AI capability improvements would impact the economy and labor market as organizations adopt AI, which could have a substantially longer timeline.

Les auteurs étudient le lien entre traits de personnalité et productivité en télétravail. À partir d’une enquête conduite en Lettonie en 2021, ils mesurent les traits de personnalité du modèle des «cinq grands facteurs de personnalité» chez plus de 1 700 personnes qui ont récemment télétravaillé. Ils constatent un lien positif entre la conscience et la productivité en télétravail. La conscience et l’ouverture sont en outre positivement associées à la volonté de continuer de travailler à distance après la pandémie. Les employeurs favorables au télétravail devraient donc être attirants pour les salariés présentant ces attributs. En revanche, le lien entre extraversion et préférence pour le télétravail est négatif. L’article montre qu’une politique uniforme a peu de chances de maximiser la productivité des entreprises et la satisfaction des salariés.

Rima N. Suleymanova

The article examines the experience of implementing the national policy in the USSR during the Great Patriotic War. In the context of modern instability in the world, complex interethnic and interstate relations, the historical experience of finding ways to cooperate, solving important issues in this area at the regional level can be very much in demand. When developing state programs and strategies, it will allow taking into account not only the useful experience accumulated in the past, but also to avoid repeating mistakes. The war made considerable adjustments to the functioning of the party and state apparatus, the activities of public organizations, the USSR citizens’ lifestyle and spirits. Using documentary sources and the case of the Bashkir ASSR, one of the country regions, the author has carried out an analysis of the socio-political and religious-cultural life of the republic, changes in the national policy and interethnic relations. The paper demonstrates the increased importance of ideological and political education of the population during the war. The author points out new trends in the activities of public and religious organizations and in the local authorities’ work on the national issue in the following areas: in personnel training; in the regulation of interethnic relations, taking into account the arrival of the evacuated population and labor-mobilized citizens of different nationalities from many regions of the country; and activities aimed at lifting up the spirits among the intelligentsia. The national policy pursued in the republic during the period under review is characterized by complexity and contradiction. That was manifested in the fluctuation from weakening at the beginning of the war, and even lack of attention to many issues of national life, to a sharp change towards strict centralization and unification at the end of the war. The study draws attention to the omissions and difficulties in the implementation of the national policy.

Aleksandra PARTEKA, Dagmara NIKULIN, Joanna WOLSZCZAK-DERLACZ

Les autrices utilisent un jeu de microdonnées sur les travailleurs de 22 pays européens afin d’évaluer si les technologies influent sur le lien entre les chaînes de valeur mondiales (CVM) et les conditions de travail mesurées par les salaires et par plusieurs dimensions de la qualité de l’emploi. Elles analysent cette influence pour plusieurs types de technologies, comparant les logiciels et robots à l’intelligence artificielle. Globalement, la participation aux CVM a un lien négatif avec les salaires et (légèrement) positif avec certaines dimensions non monétaires de la qualité de l’emploi. L’utilisation des technologies numériques ne modifie pas cette relation de manière économiquement significative.

D. E. Otighi

Objective: to provide a comprehensive legal and economic analysis of the validity of robot taxation as a measure to protect the labor market under the increasing automation, taking into account the socio-economic realities of Nigeria’s developing economy.Methods: the research is based on doctrinal and comparative legal methodology. The author systematically analyzed scientific publications, legislative acts, statistical data and empirical materials related to the impact of robotics and artificial intelligence on global labor markets. Special attention was paid to studying tax policy in the field of automation in South Korea and the European Union, in order to identify universal patterns and specific features of automation regulation in various jurisdictions. Methodological tools include content analysis of regulatory documents, economic and statistical analysis of data from international organizations, and a critical analysis of doctrinal provisions regarding the prospects for robot taxation.Results: the research demonstrates the ambiguity of the robot taxation institute in the modern legal and economic system. It was found that the robot taxation may slow down the pace of automation, provide workers with time to adapt and retrain, compensate for the reduction in income tax revenues and ensure economic equity by redistributing corporate income from automation. At the same time, significant limitations of this concept were identified: the risk of inhibiting innovation, the lack of a unified legal definition of the “robot”, the threat of capital outflow and the shift of production to jurisdictions with a more favorable tax environment. In relation to Nigeria, the conclusion is that a robot tax is premature due to low automation, high structural unemployment, the dominance of the informal employment sector, and poor digital infrastructure.Scientific novelty: the work is a systematic study of the legal and economic aspects of robot taxation in the Nigerian legal system. The study is novel as it substantiates a contextual approach to determining the feasibility of a robot tax, taking into account the stage of economic development, the structure of the labor market and the degree of penetration of automation technologies. For the first time, the author formulates the concept of responsible automation for developing economies, which implies not punitive taxation, but a system of incentives combining moderate fees with investments in human capital and digital infrastructure.Practical significance: the research results are valuable for forming state policy in the field of labor automation regulation. The proposed recommendations include the reform of corporate tax codes taking into account responsible automation, the introduction of mandatory assessment of the impact of automation on employment, the creation of a system of tax incentives for companies retraining workers displaced by technology, and the formation of a multilateral platform for ethical automation management. They can be used by the legislative and executive authorities of Nigeria and other developing countries to create legal mechanisms for regulating the digital economy and protecting workers’ rights under the technological transformation.

Andrea Cremaschi, Dae-Jin Lee, Manuele Leonelli

As artificial intelligence and robotics increasingly reshape the global labor market, understanding public perceptions of these technologies becomes critical. We examine how these perceptions have evolved across Latin America, using survey data from the 2017, 2018, 2020, and 2023 waves of the Latinobarómetro. Drawing on responses from over 48,000 individuals across 16 countries, we analyze fear of job loss due to artificial intelligence and robotics. Using statistical modeling and latent class analysis, we identify key structural and ideological predictors of concern, with education level and political orientation emerging as the most consistent drivers. Our findings reveal substantial temporal and cross-country variation, with a notable peak in fear during 2018 and distinct attitudinal profiles emerging from latent segmentation. These results offer new insights into the social and structural dimensions of AI anxiety in emerging economies and contribute to a broader understanding of public attitudes toward automation beyond the Global North.

Ankolika De, Kelley Cotter, Shaheen Kanthawala et al.

Recent studies show that users often interpret social media algorithms as mystical or spiritual because of their unpredictability. This invites new questions about how such perceptions affect the content that creators create and the communities they form online. In this study, 14 creators of algorithmic conspirituality content on TikTok were interviewed to explore their interpretations and creation processes influenced by the platform's For You Page algorithm. We illustrate how creators' beliefs interact with TikTok's algorithmic mediation to reinforce and shape their spiritual or relational themes. Furthermore, we show how algorithmic conspirituality content impacts viewers, highlighting its role in generating significant emotional and affective labor for creators, stemming from complex relational dynamics inherent in this content creation. We discuss implications for design to support creators aimed at recognizing the unexpected spiritual and religious experiences algorithms prompt, as well as supporting creators in effectively managing these challenges.

Yewei Song, Lujun Li, Cedric Lothritz et al.

Low-resource languages (LRLs) lack sufficient linguistic resources and are underrepresented in benchmark datasets, resulting in persistently lower translation quality than high-resource languages, especially in privacy-sensitive and resource-limited contexts. Firstly, this study systematically evaluates state-of-the-art smaller Large Language Models in 200 languages using the FLORES-200 benchmark, highlighting persistent deficiencies and disparities in the translation of LRLs. To mitigate these limitations, we investigate knowledge distillation from large pre-trained teacher models to Small Language Models (SLMs) through supervised fine-tuning. The results show substantial improvements; for example, the translation performance of English to Luxembourgish (EN to LB), measured by the LLM-as-a-Judge score, increases from 0.36 to 0.89 in the validation set for Llama-3.2-3B. We further investigate various fine-tuning configurations and tasks to clarify the trade-offs between data scale and training efficiency, verify that the model retains its general capabilities without significant catastrophic forgetting after training, and explore the distillation benefits to other LRLs on SLMs (Khasi, Assamese, and Ukrainian). In general, this work exposes the limitations and fairness issues of current SLMs in LRL translation and systematically explores the potential of using the distillation of knowledge from large to small models, offering practical, empirically grounded recommendations to improve LRL translation systems

Shaohui Wang

This paper develops a unified game-theoretic account of how generative AI reshapes the pre-doctoral "hope-labor" market linking Principal Investigators (PIs), Research Assistants (RAs), and PhD admissions. We integrate (i) a PI-RA relational-contract stage, (ii) a task-based production technology in which AI is both substitute (automation) and complement (augmentation/leveling), and (iii) a capacity-constrained admissions tournament that converts absolute output into relative rank. The model yields four results. First, AI has a dual and thresholded effect on RA demand: when automation dominates, AI substitutes for RA labor; when augmentation dominates, small elite teams become more valuable. Second, heterogeneous PI objectives endogenously segment the RA market: quantity-maximizing PIs adopt automation and scale "project-manager" RAs, whereas quality-maximizing PIs adopt augmentation and cultivate "idea-generator" RAs. Third, a symmetric productivity shock triggers a signaling arms race: more "strong" signals flood a fixed-slot tournament, depressing the admission probability attached to any given signal and potentially lowering RA welfare despite higher productivity. Fourth, AI degrades the informational content of polished routine artifacts, creating a novel moral-hazard channel ("effort laundering") that shifts credible recommendations toward process-visible, non-automatable creative contributions. We discuss welfare and equity implications, including over-recruitment with thin mentoring, selectively misleading letters, and opaque pipelines, and outline light-touch governance (process visibility, AI-use disclosure, and limited viva/replication checks) that preserves efficiency while reducing unethical supervision and screening practices.

Stephen Meisenbacher, Svetlozar Nestorov, Peter Norlander

Data from online job postings are difficult to access and are not built in a standard or transparent manner. Data included in the standard taxonomy and occupational information database (O*NET) are updated infrequently and based on small survey samples. We adopt O*NET as a framework for building natural language processing tools that extract structured information from job postings. We publish the Job Ad Analysis Toolkit (JAAT), a collection of open-source tools built for this purpose, and demonstrate its reliability and accuracy in out-of-sample and LLM-as-a-Judge testing. We extract more than 10 billion data points from more than 155 million online job ads provided by the National Labor Exchange (NLx) Research Hub, including O*NET tasks, occupation codes, tools, and technologies, as well as wages, skills, industry, and more features. We describe the construction of a dataset of occupation, state, and industry level features aggregated by monthly active jobs from 2015 - 2025. We illustrate the potential for research and future uses in education and workforce development.

D. Zaharov, A. Lobacheva

e article analyzes the problem of personnel shortage in Russian organizations, which is proposed to be considered from two complementary positions: the microeconomic aspect — the development strategy of each specific organization, its personnel policy, features of leadership style, existing personnel and all personnel management processes that influence the selection and retention of employees in the organization, as well as macroeconomic aspects general demographic trends occurring in the state and government polic y measures in the field of demography aimed at regulating and equalizing the current demographic situation. Statistics show that there is currently no population decline in the Russian Federation. Moreover, the population is gradually increasing. The problems of personnel shortages do not lie in the decline in population figures or in the lack of necessary qualified personnel in the labor market. The problem of personnel shortage is deeper and is associated with the working conditions required by workers in the new economic and information realities. In order to implement the Decree of the President of the Russian Federation, a new national project “Demography” was developed, the objectives of which were: increasing the overall life expectancy of the population; encouraging citizens to lead a healthy lifestyle; an increase in the average fertility rate from 1.76 to 1.78 in 2024.

Parsa Mahmoudieh, Deepak Pathak, Trevor Darrell

- -

DECRETAZO. El DNU DE JAVIER MILEI (Documental de Carlos Pronzato)    El 20 de diciembre de 2023, el presidente de Argentina, Javier Milei, diez días después de su toma de posesión, anunció en la televisión nacional la firma de un Decreto de Necesidad y Urgencia (DNU) que modificará o derogará 366 leyes que regulan una amplia variedad de actividades económicas en el país, estableciendo una profunda desregulación de la economía, destruyendo las bases jurídicas del país, en los aspectos laboral, de salud, de educación y de cultura. Poco después de que se anunciara su "decreto", se enfrentó a los primeros cacerolazos en las calles del país. Entre el 22 y 28 de diciembre realizamos este documental en Buenos Aires. *** Carlos Pronzato é um cineasta, escritor, poeta, teatrólogo e ativista social, nascido na Argentina e residente no Brasil. Artista multifacético, suas obras audiovisuais, teatrais e literárias destacam-se pelo compromisso com a cultura, a memória e as lutas populares. 

Po-Hsien Wu, Chao-Lin Liu, Wei-Jie Li

We present a hybrid mechanism for recommending similar cases of labor and employment litigations. The classifier determines the similarity based on the itemized disputes of the two cases, that the courts prepared. We cluster the disputes, compute the cosine similarity between the disputes, and use the results as the features for the classification tasks. Experimental results indicate that this hybrid approach outperformed our previous system, which considered only the information about the clusters of the disputes. We replaced the disputes that were prepared by the courts with the itemized disputes that were generated by GPT-3.5 and GPT-4, and repeated the same experiments. Using the disputes generated by GPT-4 led to better results. Although our classifier did not perform as well when using the disputes that the ChatGPT generated, the results were satisfactory. Hence, we hope that the future large-language models will become practically useful.

Atmika Gorti, Manas Gaur, Aman Chadha

Large Language Models (LLMs) are prone to inheriting and amplifying societal biases embedded within their training data, potentially reinforcing harmful stereotypes related to gender, occupation, and other sensitive categories. This issue becomes particularly problematic as biased LLMs can have far-reaching consequences, leading to unfair practices and exacerbating social inequalities across various domains, such as recruitment, online content moderation, or even the criminal justice system. Although prior research has focused on detecting bias in LLMs using specialized datasets designed to highlight intrinsic biases, there has been a notable lack of investigation into how these findings correlate with authoritative datasets, such as those from the U.S. National Bureau of Labor Statistics (NBLS). To address this gap, we conduct empirical research that evaluates LLMs in a ``bias-out-of-the-box" setting, analyzing how the generated outputs compare with the distributions found in NBLS data. Furthermore, we propose a straightforward yet effective debiasing mechanism that directly incorporates NBLS instances to mitigate bias within LLMs. Our study spans seven different LLMs, including instructable, base, and mixture-of-expert models, and reveals significant levels of bias that are often overlooked by existing bias detection techniques. Importantly, our debiasing method, which does not rely on external datasets, demonstrates a substantial reduction in bias scores, highlighting the efficacy of our approach in creating fairer and more reliable LLMs.

Alexandre Marinho Pimenta

Diante do contexto atual de militarização de escolas públicas no Brasil, o artigo revisa a teoria dos Aparelhos Ideológicos de Estado de Louis Althusser e a teoria do poder disciplinar de Michel Foucault. Indica que a utilização heurística e articulada de tais autores possibilita construir os fundamentos de uma dimensão repressivo-disciplinar da educação no capitalismo. Diante de tais diretrizes analíticas, espera-se contribuir ao estudo das práticas repressivas na/da educação e das dinâmicas de dominação política, reforçadas e rearticuladas em escolas que adotam o modelo militarizado. Palavra-chave: Educação Básica; Militarização das Escolas Públicas; Aparelhos Ideológicos de Estado; Poder Disciplinar.

Viktor Domazetoski, Maryan Rizinski, Dimitar Trajanov et al.

With the ever increasing interconnectedness among countries and industries, globalization has empowered economies and promoted international trade, capital flow and labor mobility, leading to improved products and services. However, the growing interdependence has also propelled an inherent reliance on joint cooperation which has considerably influenced the complexity of global value chains (GVCs). This plays a significant role in policy decisions, raising questions about trade risks that originate from such interdependence. In this paper, we study the impact of network linkage disturbances on the output supply and input demand of countries. We model the network interconnectedness of countries according to the latest 2016 release of the World Input-Output Database (WIOD) that includes data tables for the period 2000-2014 covering 43 countries as well as a model for the Rest of the World (ROW). We assess the shock distributions across the world economy by quantifying the changes in the network linkages using sensitivity analysis. Our contribution is in the definition of a shock tensor with the purpose of evaluating the impact of link sensitivity. The shock tensor is a straightforward yet comprehensive tool that allows us to obtain ample results at various levels of granularity when combining it with aggregation operators. Our study introduces a novel methodology that enables us to acquire input and output link sensitivities for all country pairings when an economic shock initiates or concludes within a country of interest. This innovative approach also facilitates the analysis of evolving trends in these link sensitivities, providing a comprehensive understanding of the dynamics of shock propagation across the global network. Taking advantage of the time-series nature of the WIOD, our results reveal illustrative visualizations and quantative measures that characterize patterns of shock distribution and relationships among countries throughout the period from 2000 to 2014. Our methodology and results not only uncover valuable trends but also establish a structured approach to better understand the aggregate effects of shock distributions. Thus, this study could be helpful for policy makers to assess trade relationships between countries and obtain quantitative insights for making informed decisions as well as explore the overall state of the globalization as a whole.

Iman Shaker Ardekani, Mehdi Emami Meybodi

Extended AbstractPurpose: Economic growth is influenced by two factors including the accumulation of production factors and the increase in efficiency. Efficiency is one of the important issues of the economy. Although its improvement is necessary for all countries, it is of double importance in developing countries, due to the lack of superior technology and more waste of resources and production inputs. Since the industry sector constitutes a high share of the gross domestic product and is of great importance due to its previous and subsequent connections with other economic sectors, the improvement of efficiency in this sector can lead to an increase in employment, production, and income in the entire economy. The industry sector is one of the significant energy-consuming sectors in every country, and the improvement of energy efficiency as one of the important policy tools plays an essential role in the growth of the industry. Therefore, this study examines the technical efficiency and scale of energy consumption of the industries in the provinces of the country. To this end, the non-parametric method of Data Envelopment Analysis (DEA) is used. Also, the issue of targeting energy subsidies in 2019 has had a significant impact on the cost of energy consumption in the industrial sector of the country. On this basis, the present study analyzes the technical efficiency of energy in the years after the implementation of this law. Considering the difference in the share of energy consumption according to the type of energy carriers at the industry level, we have considered four leading energy carriers for the calculation of efficiency.Methodology: The research steps include defining an appropriate model for calculating technical efficiency, determining the type of DEA model in terms of input or output-oriented and the type of efficiency with regard to scale, collecting, calculating the technical efficiency and the scale of the industrial sector of each province, and finally analyzing the results. Data envelopment analysis (DEA) is a non-parametric linear programming method for evaluating the efficiency of decision-making units (DMU). The main advantage of this method compared to parametric methods such as the stochastic boundary function is that the shape of the distribution function and the production relations do not create a limit for it. In addition to technical efficiency, scale efficiency can be obtained for all units by calculating the ratio of technical efficiency in the state of constant efficiency to technical efficiency in the state of variable efficiency. The researchers used the input-oriented multi-stage DEA model with six inputs and one output to determine the technical efficiency and the energy consumption scale efficiency of the industrial sector in 31 provinces. In this model, the variables of labor force (people), formation of real fixed capital (million rials), consumption of natural gas, diesel, fuel oil and electricity (barrels equivalent of crude oil) in the industry sector of each province are the  inputs, and the actual output of the industry sector (million rials) in each province is considered as the output.Findings and Discussion: The results of technical and scale efficiency scores of industries in the provinces analyzed with the multi-stage DEA model during the years 2011-2019 show that, in 2011 (the beginning of the subsidies targeting), only active manufacturing industries in the provinces of Isfahan, Ilam, Bushehr, Tehran, Khorasan Razavi, North Khorasan, Khuzestan, Sistan and Baluchistan, Kermanshah, Kohgiluyeh and Boyer Ahmad, Gilan, Markazi, and Hormozgan were technically efficient. The inefficiency was higher in Lorestan, Golestan, Yazd, and West Azerbaijan provinces. Also, over time, when the effect of the increase in the price of energy carriers became more evident in the subsidy targeting law, the technical efficiency was relatively improved in most of the provinces. The noteworthy point here is that, among the efficient provinces, Tehran, Bushehr, Khuzestan, and Hormozgan have had the highest share of energy consumption. Ilam, Sistan and Baluchistan, Kahgiluyeh and Boyar Ahmad provinces are among the provinces with the lowest energy shares. The industrial province of Yazd, despite having being the fourth place in energy consumption in the industry sector (about 7 percent share) after West Azarbaijan Province, has the lowest average technical efficiency score. This shows that, in this province, planning is required for extensive changes in various sectors of industry in order to increase the efficiency in Energy consumption. The evaluation of the efficiency of the scale of the industries in the country's provinces shows that, in 2014 and the beginning of the law of targeting the subsidies, the industry sector of the provinces of Isfahan, Bushehr, Tehran, Sistan and Baluchistan, Fars, Kermanshah, Gilan, Markazi, Hormozgan worked on an optimal scale, but the other provinces had inefficient scales. Among the inefficient provinces, the intensity of scale inefficiency in Ilam Province was higher than that in the other provinces, which indicates that the size of its production organization is not optimal and it can move towards an efficient scale by changing the size. This is despite the fact that, after the year 2019, a relative improvement in the scale efficiency score occurred for most of the provinces. In the last year, active manufacturing industries in the provinces of Ardabil, Alborz, Bushehr, Tehran, Khuzestan, Sistan and Baluchistan, Kurdistan, Kermanshah, Markazi, Hormozgan and Yazdbenefited from the efficiency of the scale.Conclusion and Policy Implications: In DEA models, for each inefficient unit, an efficient unit or a combination of two or more efficient units is introduced as a reference unit. In this regard, each inefficient unit should be compared with an efficient unit to reach the efficiency limit. Therefore, the reference unit should be similar in size and structure to the inefficient units that measure it. In this regard, the seven provinces of Ilam, Bushehr, Tehran, Khuzestan, Sistan and Baluchistan, Kohgiluyeh Boyer Ahmad and Hormozgan are considered as reference units for the other provinces of the same level in terms of the size and structure of the industry to improve efficiency. In terms of policy-making, it is suggested that the technical efficiency score of the provinces be supported by the government and its projects for the industry sector through the allocation of low-interest loans, industrial subsidies, tax exemptions, etc.

Tyna Eloundou, Sam Manning, Pamela Mishkin et al.

We investigate the potential implications of large language models (LLMs), such as Generative Pre-trained Transformers (GPTs), on the U.S. labor market, focusing on the increased capabilities arising from LLM-powered software compared to LLMs on their own. Using a new rubric, we assess occupations based on their alignment with LLM capabilities, integrating both human expertise and GPT-4 classifications. Our findings reveal that around 80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of LLMs, while approximately 19% of workers may see at least 50% of their tasks impacted. We do not make predictions about the development or adoption timeline of such LLMs. The projected effects span all wage levels, with higher-income jobs potentially facing greater exposure to LLM capabilities and LLM-powered software. Significantly, these impacts are not restricted to industries with higher recent productivity growth. Our analysis suggests that, with access to an LLM, about 15% of all worker tasks in the US could be completed significantly faster at the same level of quality. When incorporating software and tooling built on top of LLMs, this share increases to between 47 and 56% of all tasks. This finding implies that LLM-powered software will have a substantial effect on scaling the economic impacts of the underlying models. We conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications.