신흥국 산업인력 수요전망 방법론 연구: 직업교육 ODA 사업의 효율화 방안을 중심으로 (Approaches to Forecasting Labor Demand in Developing Countries and Their Implications for Korea’s ODA)

Abstract

Korean Abstract: 신흥국의 산업인력 양성을 지원하는 직업교육훈련 ODA 사업의 내실화를 다지기 위해서는 교육과정 설계 등 사업계획 수립 및 집행에 앞서 어떤 산업과 직종에서 산업인력 수요가 어느 정도로 예상되는지 등에 대한 합리적인 진단이 중요하다. 오늘날 신흥국들은 국가별로 차이는 있지만 경제구조 전환 과정에 있고, 노동시장을 둘러싼 제반 환경이 복잡해지고 있어 기존의 노동시장 수요예측 방법만으로는 예측력을 높이기 어렵다. 본 연구는 신흥국의 특성에 맞는 정량적 기법과 정성적 기법의 결합을 시도하기 위하여 베트남을 대상으로 산업 및 직업별 인력수요 전망 분석 방법론을 시범 적용했으며, 나아가 글로벌 가치사슬(GVC) 분석과 빅데이터(big data) 분석방법론을 다루었다.

English Abstract: Technical and vocational education and training (TVET) programs are a critical area within the official development assistance (ODA) mandate of Korea. Korea is implementing TVET projects in a wide range of regions, including Southeast Asia, Central Asia, Latin America and the Caribbean, the Middle East, and Africa. Considering the increasing demand for TVET from developing countries, Korea is expected to continue and expand its TVET projects. Korea has not been conducting systematic labor market forecast in the preparation and implementation of TVET projects. Developing systematic labor forecast methods and increasing predictability of labor demand should precede actual hardware supports, including the construction of job training centers and distribution of training materials as well as curriculum development, in order to increase the efficiency of TVET projects. However, the labor demand forecasts conducted by Korea up to now largely fall short of these expectations. This study aims to develop systematical labor forecast methods thereby contributing to increasing the efficiency of TVET ODA. Towards this, the study suggests a labor forecast method that considers characteristics of developing countries, and applies the method to Vietnam as a pilot application. Based on the results, this study estimates Vietnam’s future labor demand by industry and by occupation. In addition, this study introduces alternative labor market projection methods that can possibly deal with new changes observed in the labor markets of emerging countries. Chapter 2 reviews Korean TVET ODA projects and evaluates the demand side of the projects, for instance whether any labor market forecasts had been conducted for these projects or not, and the scope and method of labor market forecast. We examine project proposals, project validity studies, and evaluation studies for evaluating how much the demand side was considered in the process of implementing TVET projects. In Chapter 3, as a precursor to developing an alternative labor projection method, we introduce labor forecast models used in developing and advanced countries and compare their characteristics. Labor forecast studies started as the issue of effective human resource allocation emerged as a national task for many countries after World War Ⅱ. Accurate macroeconomic projection is critical for labor forecast, because labor demand is basically derived demand generated as the aggregate demand of economic increases. Recently, econometric methods based on the general equilibrium model are largely used in order to include inter-related effects of sectors, by using population, macroeconomic, industrial, and labor time series data. As the model becomes larger, several countries divide the economy into several blocks (e.g. macroeconomy, international economy, labor market, etc.) and focus on the dynamic estimation in each block. Furthermore, qualitative methods are used to take account for the impact of emergence of new industries and decline of existing industries due to technological innovation and international trade environmental change. In Korea, the Korea Employment Information Service was established to perform projections of labor demand and supply starting from 2006. This labor forecast reflects changes in population, society, economy, and industry structure. The projection model is largely composed of a section dealing with aggregate labor supply and demand forecast, and a forecast of new labor force. The section on aggregate labor supply and demand forecast estimates the stock of labor supply and demand based on the composition and trends in the labor force, and industrial growth prospects. The section on new labor force forecast estimates the flow of labor supply and demand and the supply-demand gap for a certain time period and looks at how new labor supply and demand change by education, major, and technology. The labor supply forecast estimates the labor force based on population prospects using census data and economic participation rate projections using labor force survey data. The labor demand forecast estimates real value-added based on industry prospects using an industry-macro econometric model with national income account data, and labor demand forecast by industry by multiplying employment coefficients earned from labor and production time series data by industry. This is used for estimating labor demand forecast by industry and occupation in addition to an estimated industry-occupation labor demand matrix. Finally the recursive structure of the model satisfies the matched aggregate labor supply and demand at the natural unemployment rate level. Labor forecast methods used in most advanced countries are similar to the Korean model. The US Bureau of Labor Statistics model uses segmented data by industry and occupation, while the Dutch ROA model is characterized by segmenting and forecasting job prospects by sector or level of education in the underlying labor markets. In most developing countries, the lack of statistics data or forecasting capacity and experience makes it difficult to implement a systematic labor market forecast. Some countries are developing methodologies to build statistical information and forecasts for industrial labor demand with the support of international organizations such as the International Labor Organization or donor countries. However, these projects are likely to end as one-time events while performing forecasts on industrial labor demand involves huge expense and experts. This makes it difficult to systemize and maintain the industrial labor demand forecasting system independently. Quantitative and qualitative analysis have different strengths and weaknesses. It is necessary to increase prediction power through a “combination” of the analytical methods rather than relying on any one method. Combining analytical methods needs more time and cost than using one method, but various methods can be utilized to increase the reliability of predictions. In Chapter 4, we applied our labor forecasting methodology to Vietnam, an exemplary emerging country, to estimate future labor demand by industry and occupation, as a pilot analysis that combines quantitative and qualitative analyses for emerging countries. Vietnam was chosen for the pilot because of the following reasons. The level of labor market statistics is relatively high compared to other emerging economies, while the manufacturing industry, a key industry in vocational education ODA projects, accounts for a high proportion in the economy. Also, Vietnam is the largest recipient of TVET assistance from Korea. First of all, we use data from the General Statistics Office of Vietnam to forecast labor demand for 20 major classification industries and nine major classification occupations. Based on the forecasting methodology used in Korea, the outlook for industrial and occupational classification was implemented, but due to the characteristics of emerging countries that have short time series data and less stable acroeconomic prospects, we could only perform a mid-term forecast for the period of 2019 to 2024. Our results indicate a significant trend difference by industry though the total number of employees is expected toincrease by 0.5% annually on average by 2024. Employment in the manufacturing sector is expected to increase by 2.4% per year, while employment in primary industries such as the agricultural and mining sectors is expected to decrease by 0.8% to 3.1% per year. By occupation, employment in professional groups is expected to increase by 5.3% per year, while elementary occupations and skilled agricultural, forestry and fishery workers are expected to decrease by 1.9% to 3.5% per year. In addition, according to the purpose of this study, we implemented a labor demand forecast for detailed industries and occupations within the manufacturing industry, which is at the core of vocational training. Due to the lack of detailed statistics on the manufacturing industry within the data from the General Statistics Office of Vietnam, we chose to use data taken from the UNIDO Industrial Statistics (INDSTAT) Database 4 on value-added and number of employees by industries in the manufacturing industry. Total employment in the manufacturing industry is expected to increase by 1.7% annually during the 2017-2024 period. Employment in the manufacture of communication equipment (Industry 263 in ISIC Rev. 4) increased by 47.3% and 35.8% annually for 2007-2012 and 2013-2016, respectively, but the annual growth rate of employment in this sector is expected to adjust to 3.9% annually during 2017-2024. Currently, traditional quantitative forecasting methods can be applied only to 161 minor groups (3 digits) of industries in Vietnam, because unit-level (4 digits) statistics have not been fully established. Accordingly, using non-traditional methods is necessary in labor forecast for further detail industries at unit level. We applied a hybrid method of quantitative and qualitative approaches to the manufacture of wireless communication equipment, one of the most prominent sectors recently in the Vietnamese manufacturing industry, and given that TVET projects mostly target the manufacturing sector, this study chose model cases among the manufacturing industries. Import and export statistics, enterprise survey, and interview of Vietnamese government officials and experts on Vietnamese industry are conjointly used in the labor forecast of wireless communication equipment, based on the