Review
Focus on inflation in 2022
The research conducted by KOF relates closely to the economic policy issues affecting society. Acting as a bridge between research and society lies at the heart of the institute’s mission statement and therefore forms the basis of many research projects. In addition to the war in Ukraine, the main economic topics in the public debate last year were surging inflation and the much-discussed energy crisis.
As well as offering its expertise on these subjects in the media, KOF launched various projects focusing on the issue of inflation. The institute’s Business Tendency Surveys, for example, included new quantitative questions on the participating firms’ forecasts for wages and inflation. What is interesting here is that this provides a direct assessment of what sort of wage growth firms are expecting in their operations and what their price projections look like. These developments were supported by external individuals at the universities of Basel and Lüneburg and by the Swiss National Bank.
The results are made publicly available as experimental statistics. Building on this project, KOF conducted an experimental survey to answer the question of whether and, if so, how inflation expectations affect firms’ price-setting and wage-setting behaviour. The findings obtained from the survey can provide general information on how price pressures spread throughout an economy.
Shortage of skilled labour and monitoring of the labour market throughout the Swiss economy
A further frequent topic of public debate was the shortage of skilled labour. In addition to being analysed in the media, it is also the subject of research projects at KOF. Working as part of a external pageproject funded by the Swiss Employers Confederation (in German), KOF researchers collaborated with an external partner to devise a new method of characterising the shortage of skilled workers in Switzerland in various sectors, professions and specialisms in a comprehensive and detailed way. For this purpose they used online data that showed how long virtually all jobs advertised on the internet remain available. It was the first time that this data had been used for this kind of evaluation. It enables researchers to identify job profiles that are especially difficult to recruit. The final report was published in February 2023.
Another project that examines the issue of promoting skills in science, technology, engineering and maths (STEM) addresses an early stage of the subsequent shortage of skilled labour. The aim of the cross-institutional ‘Edumap’ project is to promote the teaching of STEM subjects in secondary schools and to support these schools by providing them with data analysis to encourage pupils to study STEM subjects. The project is being conducted as a collaboration with the Department of Economics at the University of Zurich, the Assistant Professor of the Economics of Child and Youth Development at the University of Zurich and the Executive Boards of ETH and the University of Zurich.
This involved processing data on all of the students at the University of Zurich and ETH Zurich since 2010. The project also analysed various case studies to identify what individual and school-related factors play a role in pupils’ decision to study a STEM subject at one of these two universities. The relevant findings were made available to schools online at the end of 2022. This offering attracted considerable interest and, at the same time, closes a knowledge gap. As initial evaluations show, school heads have no systematic information on what subjects their school-leavers choose to study at university.
KOF is conducting a project entitled ‘What Workers Want: Determinants and Implications of Job Search Strategies on an Online Job Platform’ in collaboration with the Chair of Applied Economics and Econometrics at HEC Lausanne external pageas part of the ‘Digital Transformation’ national research programme (NFP 77) under the auspices of the Swiss National Fund (SNF). New kinds of data from a public recruitment platform are used to investigate how unemployed individuals search for jobs online.
This project aims to improve understanding of jobseekers’ success factors as well as trends in the Swiss labour market and to optimise how employment offices find jobs for individuals. One of the first milestones of this project was the launch of a high-frequency Job Tracker, which monitors the number of jobs advertised online in Switzerland over time. The Swiss Job Tracker is updated once a week and therefore enables the Swiss labour market to be monitored in real time. https://kof.ethz.ch/en/forecasts-and-indicators/indicators/swiss-job-tracker.html.
Digital technology and innovation in the Swiss economy
A further NFP 77 project entitled ‘Digital transformation: how it changes organizations, performance, and markets – a multi-level analysis’, which is being conducted in cooperation with the Chair of Strategic Management and Innovation at ETH Zurich, is investigating how digital transformation impacts on the performance of firms in Switzerland and on market dynamics. The aim here is to develop a representative database and identify the implications for policymakers and strategists.
Innovation drives development and growth. That is why, for the past three decades or more, KOF has been surveying Swiss firms’ level of innovation and, for some time now, the amount of digital technology that they use as well, having been commissioned to do so by Switzerland’s State Secretariat for Education, Research and Innovation (SERI). The latest innovation survey, covering the period from 2018 to 2020, was published in 2022 (see the case study on page 10 for further details).
Traditional data gathering methods such as surveys are not always appropriate for capturing new technologies in a timely and comprehensive way. The ‘web-based innovation metrics’ project aims to use online data to devise networking and innovation indicators in the field of artificial intelligence. These indicators enable the latest innovation activities and trends throughout the Swiss economy to be captured. The project, which is being conducted in collaboration with the University of Hohenheim and Istari.ai, is due to be completed in 2023.
Refinement of forecasting methods
Good-quality estimates and forecasts are crucial so that decision-makers from the fields of business, politics and society can make rational and well-informed decisions. What is important here is not only long-term forecasts that capture structural changes in the economy and society but also timely short-term estimates and forecasts that can identify and quantify economic turnarounds at an early stage.
New methods such as mixed-frequency data models (MIDAS), artificial intelligence (AI) approaches and dynamic factor models now enable large quantities of data to be analysed virtually in real time so that short-term movements within key economic data can be identified. This was very much evident during the COVID-19 pandemic, when the economic and financial situation was changing almost daily, if not by the hour initially. By using various methods, it is possible to combine model forecasts in order to improve forecasting accuracy. Various projects at KOF aim to use exactly these new methods to develop new models for short- and longer-term forecasts.
One of these projects is external pageKOF’s Nowcasting Lab, which is a real-time testing platform for forecasting current-quarter GDP by using previously available and higher-frequency data. These models are updated daily for a number of countries based on large quantities of data and are published online. KOF’s Nowcasting Lab started its second year successfully in 2022, acquiring the Directorate‑General for Economic and Financial Affairs (DG ECFIN) of the European Commission as a collaboration partner. In addition, further countries were integrated into the Nowcasting Lab. Portugal was included in the forecasts on behalf of the Portuguese finance ministry. Bulgaria, Romania and the United States were also added.
When forecasting the Swiss economy, KOF usually used its ‘macro-model’, which was completely overhauled and converted into a Bayesian estimation method, for which a new type of numerical procedure was devised. This approach helps to provide a better and more intuitive assessment of the uncertainty inherent in any forecast. The new ‘KOF KoMa’ forecasting model is due to be used for the first time in 2023. There are also plans to make the new model’s Bayesian estimation method publicly available.
A team is developing financial forecasting models over various time horizons for and in collaboration with Switzerland’s Federal Finance Administration (FFA). The ‘Fiscal Nowcasting’ project is devising ways of providing financial statistics data on a quarterly basis. Such data has traditionally only appeared once a year. However, the recent economic crises in particular have increased the need for quarterly data.
The ‘Macroeconomic Fiscal Forecasting’ project is designed to produce short- and medium-term forecasts of the key aggregates of public budget data, which are consistent with the development of macroeconomic indicators. This approach is based on the Bayesian vector autoregressions (BVARs) being developed at the institute. The aim of this sub-project is to supplement KOF’s models with additional fiscal variables that are relevant to fiscal policy and the production of government financial statistics.