76211 A New Jobs Data Tool: Introducing BuDDy – a Business Diagnostics and Dynamics Tool China’s have an inclusive growth agenda, as they are faced Dino Merotto and Jessica Boccardo with disparities in job opportunities and wide inequality. The World Bank’s World Development Report 2012 puts Key Messages1 jobs at center stage. Policy makers are thus in critical need of data, facts, and advice on what to do to help stimulate  The BuDDy tool uses formal sector business data jobs for prosperity in their countries. that governments already collect to analyze patterns and trends in employment and diagnose constraints Clearly, to make informed decisions about policy priorities to growth and job creation. for job creation, the necessary information must be  BuDDy gives governments the understanding of available. The most often used and commonly available business dynamics needed to develop policies that source of information comes from the World Bank’s help businesses create jobs. Household Labor Force Surveys (LFS), which are linked to the living standards surveys (LSS). These surveys, which  BuDDy quickly and robustly identifies the types of look at workers and not firms, provide insights on how jobs firms that are growing, hiring, investing, raising help households rise out of poverty. The Business productivity, and raising real wages, and does this at Diagnostics and Dynamics Tool (BuDDy) complements this the national or regional level, or by product. analysis with detailed information from formal sector firms,  BuDDy is simple and adaptable; it has been giving demand-side insights into how, why, and where developed with varying data sets, and can been “good jobs� are created and lost in the economy. BuDDY linked to spatial information, trade data, and was created by staff originally working in the Africa (AFR) household data sets. and Europe and Central Asia (ECA) Poverty Reduction and Economic Management (PREM) units. Currently, the Introduction PREM Anchor is expanding this work so that World Bank teams can help meet this demand for precise information on Policy makers around the world want more and better job creation country-by-country. jobs for their citizens. Both emerging and industrial countries that lost jobs in the global crisis now confront How does BuDDy Work? jobless recoveries. The “Arab Spring� of 2011 brought to light the imperative of job creation for predominantly young BuDDy processes readily available data, applying standard populations. Demographics suggest that urbanization and diagnostics from the growth literature. In most countries, underemployment in the next generation of youth could be the data to help understand job creation and growth have even more pronounced in sub-Saharan Africa. In Eastern already been collected, but are not being used. National Europe, out-migration and aging has caused labor markets accounts surveys that are collected by statistical agencies to to shrink and participation to fall, and the lack of estimate GDP on a value-added basis take operating competitiveness has led businesses to shed workers, adding statements from firms in order to estimate sales, costs of to the out-migration. Even faster growing economies like sales, profits, labor costs, and other indirect costs for the economy. To generate an investment series for the national 1 The authors are grateful to Indermit Gill, ECA Chief Economist, accounts, statisticians also collect information on trends in Yvonne Tsikata, ECA PREM Director, and Benu Bidani, Sector gross fixed capital formation from businesses. These Manager, ECA PREM; to Jeff Lewis, PRMED Director, and Shan Gooptu, PRMED Sector Manager, for supporting the development of the surveys and the census from which samples are drawn BuDDy; and to the ECA Innovation Grant for funding. typically collect information on how many workers a firm ECA Knowledge Brief employs. All of these sources of data are combined across Getting a Good Diagnosis years in a BuDDy. Ideally, this data will, in time, be standardized with growth analysis in mind, just as Facts are collected using BuDDy to enhance the analyst’s household LSS are increasingly becoming standardized. For diagnosis of the economy. A good diagnosis starts with now, there is a treasure of data across countries to be mined whether the observed symptoms of growth and job creation and explored. As a tool for making the most of existing reflect those of a healthy economy. With BuDDy, some data, BuDDy costs very little and offers much. standard tables and charts are made that allow the user to identify facts about growth dynamics. Is there churning; i.e., The nature of the data used for BuDDy means it is country- do efficient firms grow, and do inefficient firms shrink? Is specific. Some statistical agencies collect business registers there entry and exit, and do entrants tend to have higher (censuses) that record basic data for every registered firm. than average productivity? Do growing firms hire, and do Where firms are matched over time periods, this allows for shrinking firms fire staff? Are profits and job creation a comprehensive analysis of entrants, exits, growth, and significantly higher for efficient and more productive firms? sector concentration. Some agencies conduct annual Are firms with rising productivity hiring and investing? Is surveys, following a few of the same firms over time, the spatial pattern of new business location and growth as usually providing more detailed operating statements and expected? cost data, or detailed financial statements and balance sheets. Some collect the bare minimum to generate GDP BuDDy is designed to be simple and flexible. Users can value added growth for the national accounts. arrange data using different combinations of intrinsic characteristics and conditions, or choose (using filters) any Data on each business are used to generate standard labels specific firm or product characteristics. For instance, users and variables for analytic insights. All surveys include of Moldova’s BuDDy can filter the data to focus only on intrinsic characteristics about a business: where it is efficient firms in manufacturing; they can then drill down to located, who owns it, what product(s) it produces, how old look at job losses in any given district by firm ownership and how large it is, etc. Surveys also include current data on and age. Alternatively, a user could investigate the returns the intrinsic conditions of the business: its sales, costs, to and costs of capital for efficient and inefficient firms, and profits, employment numbers, etc. When available over then check whether this pattern holds across firm time, these intrinsic data are used to generate derived ownership, age, and sectors. The tool comes with a user conditions such as growth rates, productivity, profitability manual, and Bank teams are providing growth diagnostic ratios, returns to capital, costs of capital, average wages, training on how to develop and use it for counterparts in capital/labor ratios, and so forth. A combination of countries that have made their data available. characteristics and derived conditions is then used to label and differentiate businesses with derived characteristics — The firm surveys used to create BuDDy are generally much for instance, growing vs. shrinking firms, efficient vs. larger than those collected in Business Environment and inefficient, productive vs. unproductive, profitable vs. loss- Enterprise Performance (BEEPs) and International making, high vs. low return, old vs. new, entrants vs. Investment Climate (ICA) surveys. Moldova’s BuDDy, for established firms. Businesses are thus labeled based on their example, created a panel of 17,000 firms from 2004 to product characteristics, that is, producers are differentiated 2010; Ukraine’s panel is made up of 37,000 firms, and the based on whether they produce traded or nontraded goods, data panel for Romania is 102,000. In Uganda, there are two or products that are “monopolistic� as defined by their censuses a decade apart; the 2001 Census contains 166,000 concentration coefficient, or “competitive.� Finally, other firms, the 2010–11 Census has 450,000, and the Uganda product characteristics are incorporated, such as a firm’s national accounts survey for 2001 draws a sample of 4,400 technology content or export sophistication. firms from the census. Armenia’s national accounts survey of around 6,000 firms creates a panel of 402 firms over five These dynamic conditions and characteristics are then used years; though it covers fewer firms, the survey collects by the BuDDy tool to understand inclusive growth. Policy business costs in great detail. makers don’t just need to know the average growth rate of employment in an economy, they need to know the patterns, For the ECA BEEPs BuDDy tool, there are 22 countries for conditions, and characteristics of the firms that are hiring which panel data are available for 2004 and 2007 in surveys and those that are not. BuDDy allows users to contrast averaging roughly 400 firms per country. The surveys hiring firms from firing firms. Are hiring firms investing? contain rich information on investment climate constraints, Are their sales rising? Are they more efficient and but the lack of financial capital and data on gross fixed profitable? Is their productivity higher than that of firing capital formation means that returns to capital and firms? Do they pay higher wages if it is? production functions cannot be derived. Our pilot work ECA Knowledge Brief suggests that despite the variety in data contexts, BuDDy lagging regions in relation to productivity or value added has a lot to offer policy makers. (figure 3). What Can BuDDy Do for You? 2 Figure 2. Romania: Job Creation and Destruction in Manufacturing Before and After the Global Crisis Trend analysis: Who is creating jobs, in which sectors and where? With the click of a button, BuDDy provides an accurate view of which types of firms created jobs and where. And it can generate facts about the drivers of job creation. Standard tables show trends in employment and labor productivity broken down by location, ownership, firm size, and sector or product. Figure 1, for example, shows the trends in employment, wages, and value added per employee for firms in Ukraine’s data panel. What stands out is that Ukraine exhibited growth in value added with job destruction and rising wages prior to the global crisis. Since then, however, wages and employment have been hard hit. Source: BuDDy team calculations (Kanematsu 2012) based on the Romania Figure 1. Trends in Employment, Wages, and Value Added in Amadeus database. Bubble size is employment in 2009. Ukraine Inclusive growth diagnostics: Do efficient firms grow, and do growing firms hire? Inclusive economic growth requires that they do. Otherwise the dividends of growth go not to the lower and middle classes but as returns to investors. BuDDy can be used to quickly check whether firms that are investing and expanding are the most productive in terms of output per worker. If they are, do they hire additional workers as they grow? Figure 4 shows that whereas efficient firms in Moldova do grow faster than inefficient firms, they do not (on average) hire as they grow. This pattern holds across the whole country. In the cities, the difference in hiring and firing between efficient and Source: BuDDy team calculations (Kongchanagul 2012) based on the Ukraine inefficient firms is negligible. Amadeus database. Figure 3. Average Wage/Worker by Country, Romania (2006) How was job creation affected by the crisis, and who 7000 Ilf was worst affected? Figure 2 shows the growth rates of ov manufacturing employment for a panel of Romanian firms before and after the crisis. At the level of products, it is Hunedo 6000 Gal possible to distinguish some that remained robust during the ara ati Si crisis, and others that were much more vulnerable to bi Ti Praho Tulc Mur Consta u mi Wage volatility in employment. Fabricated metal and food nta va 5000 ea es s Average manufacturing stand out as both robust and important Dambo Valc sectors in the job market in Romania. vita ea Bras C B ov l 4000 ac Are there regional patterns to employment growth or u u Mehedi wages? BuDDy allows us to consider levels and changes in nti j labor productivity, wages, and employment across regions 3000 in a country. This can be done by sector and by firm type, 0 1000 2000 3000 for instance, to check whether there may be persistent VA (000s) 2 This section uses work of the broader BuDDy team: Apirat 2 Kongchanagul, Kosuke Kanematsu, and Tao Huang, (Columbia R =62% University interns), and Charles Udomsaph (Georgetown University), p value=0 with support from Erwin Tiongson (ECSPE), Rachel Sebudde (Uganda), Y= 3517.56 + 1.02X Iaroslav Baclajanschi (ECSPE), Supo Olusi (ECSPE), and Anca Rusu (Moldova). Source: Kanematsu, 2012, based on Romania Amadeus database. ECA Knowledge Brief Do firms with higher labor productivity pay higher but there is a lot of churning, with many firms shrinking and wages? A second ingredient of inclusive growth is that many firms growing and very large firms – as research by these growing and hiring productive firms pay higher wages the World Bank has shown is also the case for Latin than less productive firms. Where they do, the economy America – predominantly shrinking. Most likely these are benefits, as workers are attracted to the higher wages and firms restructured under outward-oriented, private sector- labor shifts to more productive employers. Figure 5 shows led reform programs. that for Moldova, value added productivity and wage Figure 6. In Uganda Micro and Small Enterprises Grow on growth is highest among firms with some form of foreign Average – With a Lot of Churning ownership. These firms are few in number, but their wage Evidence of Churning Amongst Small and per worker seems to rise faster than all other business Micro Firms in Uganda 2001-2010 ownership groups, except state ownership. Figure 4. Efficient firms in Moldova do invest, especially 5337 Sum of Change In Employment outside the cities; but they don’t hire as they grow Count of businessid01 Change In Capital Change in Labor 2212 1927 1574 5,000 25,000,000 149 287 20,000,000 - Chisinau Balti Other All 15,000,000 (5,000) Growing Shrinking Growing Shrinking -942 10,000,000 Micro Small (10,000) 5,000,000 -2342 - (15,000) Source: Authors’ calculations, BuDDy database, and national data sources. Inefficient Chisinau Balti Other All (5,000,000) (20,000) (10,000,000) Efficient The Path Ahead (25,000) Inefficient (15,000,000) (20,000,000) Efficient (30,000) In the future, the BuDDy team in the Economic Policy and (25,000,000) (35,000) Debt Department (PRMED) will continue to improve this Source: Merotto and Boccardo based on Moldova Financial Statements survey. tool, working along four dimensions:  Build BuDDy in more countries Figure 5. Foreign firms in Moldova have higher productivity  Develop a program of capacity building for policy and wage growth makers and analysts Productivity is highest and is growing Wages can rise faster in foreign firms  Refine standard diagnostic tests based on the knowledge most in foreign owned companies with rising productivity of what are healthy and unhealthy symptoms in Change In Wages, Health, Pensions Per Worker business dynamics and job creation 1,400 Value Added Per Worker 250  Develop policy-relevant “syndromes� around these 1,200 200 symptoms, and catalogue policies that have helped 1,000 2004 2009 150 business growth and jobs in these environments  Showcase how BuDDy is being used for precise, faster, 800 600 100 400 deeper, and more collaborative policy advice 50 200 To learn more about the BuDDy initiative, please contact - Foreign JV Municipal PPP Private State - Foreign JV Municipal PPP Private State PRMED staff, or email directly dmerotto@worldbank.org. About the Authors Source: Merotto and Boccardo based on Moldova Financial Statements survey. Dino Merotto is a Lead Economist in PRMED, and Growth dynamics: do small firms grow and hire? In previously worked in AFR and ECA PREM. Jessica low-income countries (LICs), microenterprises and Boccardo is a Consultant and Researcher at New York informality dominate the job market. Growth and University, New York. urbanization are leading to the rapid entry of many About the Broader Team thousands of small firms into the economy. Though Charles Udomsaph is Visiting Assistant Professor, business data are generally less well developed in LICs, it is Georgetown University. Apirat Kongchanagul, Kosuke possible in some countries to quickly get the facts about Kanematsu, and Tao Huang, now Consultants were business dynamics using the BuDDy tool. Figure 6 shows Columbia University interns. Erwin Tiongson is a Senior that in Uganda, micro and small firms do grow over time, Economist in ECA PREM. “ECA Knowledge Brief� is a regular series of notes highlighting recent analyses, good practices, and lessons learned from the development work program of the World Bank’s Europe and Central Asia Region http://www.worldbank.org/eca