Combining job mobility patterns and vacancy data to better measure labour market opportunities and skill mismatch

新冠肺炎疫情给劳动力市场带来了前所未有的冲击。酒店业等行业的就业机会一夜之间消失了，而其他行业（如驾驶和医疗保健）的就业机会却在增加。这种彻底的经济重组需要一种新的方法来衡量个人面临的劳动力市场前景，这些个人通过转移到其他可获得的工作领域，抓住了他们适应冲击的巨大差异的能力。受此启发，我们开发了一种新的针对工人的劳动力市场机会衡量标准。我们的衡量标准是基于从一份工作到另一份工作的历史模式，我们用它来推断一个给定的工人对不同类型工作的适合程度。结合关于职位空缺的高质量数据，这使我们能够超越劳动力市场表现的总体衡量标准，并量化不同类型工人在大流行期间面临的劳动力市场前景的重要差异（Costa Dias等人2021年，Joyce等人2022年）。虽然我们的创新是由大流行病引发的，但我们的方法的有用性要普遍和广泛得多。我们的方法将一个经济部门的劳动力市场状况与其他经济部门联系起来，这意味着它可以用来回答这样的问题：零售业的冲击如何影响送货司机的工资压力和职业发展？更广泛的劳动力市场的变化如何影响社会护理部门招聘和留住工人的能力？我们的方法进一步深入了解了工人的技能与雇主要求的技能相匹配的程度——劳动力市场“不匹配”的程度——这对生产率和经济增长有影响。例如，如果最近离开劳动力市场的老年工人回来，这在多大程度上可以缓解特定职业的劳动力短缺？在经济中，技能短缺最严重的地方是哪里？就总生产率而言，不匹配的经济成本是多少？该项目的目的是扩展和完善我们的新方法，将其建立在理论搜索和匹配模型的基础上，并展示其对实质性问题的适用性及其相对于传统方法的优势。我们将改进我们的指数，以反映不同类型求职者竞争工作所产生的机会差异，以及雇主所要求的技能差异（我们目前的指数反映了这一点）。然后，我们将通过将其应用于经验，与政策相关的问题（如上述问题）来证明其效用-特别是在确定劳动力市场机会在工人和时间上的变化，以及量化经济中劳动力市场不匹配的程度和经济成本方面。将这一创新嵌入未来研究和统计实践的一个关键部分将是传播和与他人的接触——不仅在项目完成之后，而且在项目的生产过程中，帮助我们最大限度地发挥其效用。主要目标受众将包括研究界、统计从业人员（如国家统计局）和政策制定者——因为我们创新的最终目的是更好地指导政策制定，通过对政策制定者产生影响，我们将在未来的研究中进一步突出我们创新的潜力。

The Covid pandemic brought about an unprecedented labour market shock. Job opportunities in sectors like hospitality disappeared overnight, whilst opportunities expanded in other sectors like driving and healthcare. This radical restructuring of the economy called for a new way of measuring the labour market prospects facing individuals that captured their widely-varying ability to adapt to the shock by moving into alternative lines of available work. Motivated by this, we developed a new worker-specific measure of labour market opportunities. Our measure is based on historical patterns of job-to-job moves, which we use to infer how suitable a given worker is for different types of jobs. Combined with high-quality data on vacancy postings, this allowed us to look beyond aggregate measures of labour market performance and quantify important differences in the labour market prospects facing different types of workers over the pandemic (Costa Dias et al 2021, Joyce et al 2022). While our innovation was triggered by the pandemic, the usefulness of our method is much more general and wide-ranging. Our method links the state of the labour market in one part of the economy to the rest of the economy, meaning that it could be used to answer questions like: how do shocks in retail affect wage pressures and career progression for delivery drivers? How do changes in the wider labour market affect the ability of the social care sector to recruit and retain workers? Our method further provides insights into the extent to which workers' skills align with the skills demanded by employers - the extent of labour market 'mismatch' - which has implications for productivity and economic growth. For example, if the older workers who have recently left the labour market were to come back, to what extent could this ease labour shortages in specific occupations? Where in the economy are skills shortages most acute, and what is the economic cost of mismatch in terms of aggregate productivity?The aim of this project is to extend and refine our new method, grounding it in a theoretical search and match model, and demonstrate its applicability to substantive questions and its advantages over traditional methods. We will refine our index to capture variation in opportunities arising from competition for jobs from different types of jobseekers, as well as variation in the skills demanded by employers (captured by our current index). We will then demonstrate its utility by applying it to empirical, policy-relevant questions like the ones above - in particular, in identifying the variation in labour market opportunities across workers and over time, and in quantifying the extent and economic cost of labour market mismatch in the economy. A key part of embedding this innovation within future research and statistical practice will be dissemination and engagement with others - not only after the project is done, but during its production, helping us to maximise its usefulness. The main target audiences will include the research community, statistical practitioners (e.g. ONS), and policy-makers - since the ultimate purpose of our innovation is to better guide policy-making, and through achieving impact on policy-makers we will further highlight the potential of our innovation within future research.