CHAPTER 1: INTRODUCTION
Problem Statement
In recent decades, several emerging countries have experienced significant economic growth driven by foreign direct investment (FDI) This capital inflow has led to increased local employment, the introduction of advanced equipment, and the practical application of new technologies Additionally, a notable indirect effect is the technology spillover from multinational corporations' subsidiaries to local firms, further enhancing the host countries' economic development.
Between the 1960s and 1970s, foreign direct investment (FDI) played a crucial role in the rapid economic growth of Hong Kong and Singapore This trend is mirrored in recent years by China and Vietnam, with Vietnam emerging as one of the fastest-growing economies globally, boasting an average growth rate of over 7 percent per year, largely attributed to its success in attracting substantial FDI.
The impact of foreign direct investment (FDI) inflows on the economic growth of host countries is often overlooked, yet there is significant discourse surrounding the potential for FDI to generate technology spillovers that enhance the labor productivity of domestic firms.
Foreign Direct Investment (FDI) can have varying effects on the productivity of domestic firms, with some studies, such as those by Caves (1974) and Blomstrom and Persson (1983), highlighting positive technology spillovers through competition, employee training, and demonstration effects Conversely, research by Aitken and Harrison (1999) and Kathuria (2000) suggests that FDI may negatively impact labor productivity due to market and skill stealing These mixed findings indicate that the technology spillovers associated with FDI can occur within industries, known as intra-industry or horizontal spillovers.
Research on the horizontal spillover effects from multinational corporations (MNCs) to Vietnamese firms remains limited in Vietnam However, mixed evidence has emerged from studies conducted by Le Thanh Thuy (2007), Nguyen Thi Tue Anh et al (2006), Quoc Hoi Le and Richard Pomfret (2008), Anh Ngoc Nguyen et al (2008), and Chuc D.
This research examines the horizontal technology spillover effects of foreign direct investment (FDI) on the labor productivity of domestic firms within the manufacturing sector, specifically focusing on the food processing industry as a representative low-technology sector and the electronics and mechanics industries as high-technology sectors Due to data limitations, the study does not encompass the entire manufacturing industry, unlike other studies Utilizing various panel regression models, including pooled regression, fixed effects regression model (FEM), and random effects model (REM), the study aims to uncover significant findings that can aid policymakers and businesses The results are expected to provide actionable recommendations for leveraging FDI to enhance labor productivity in both low- and high-technology industries.
Research Objectives
This research aims to examine the horizontal technology spillover effects of foreign direct investment (FDI) on the labor productivity of domestic firms in Vietnam, specifically within the low technology sector, such as food processing, and the high technology sector, including electronics and mechanics The study will concentrate on three key objectives to provide a comprehensive analysis of these effects.
(i) Whether horizontal technology spillovers through FDI affect the labour productivity of domestic firms in low technology industry as Food processing
(ii) Whether horizontal technology spillovers through FDI affect the labour productivity of domestic firms in high technology industry as Electronics and Mechanics
(iii) Recommendations for government planners and domestic firm managers to take advantage of FDI to improve the labour productivity of domestic firms.
Research Questions
This research investigates the impact of horizontal technology spillovers on the labor productivity of domestic firms in both low technology industries, specifically food processing, and high technology industries, particularly electronics and mechanics The primary aim is to explore how these spillovers influence productivity levels across different sectors.
(1) Do horizontal technology spillovers through FDI give significant impact on labour productivity of domestic firms in low technology industry as Food processing?
(2) Is labour productivity of domestic firms in high technology industry as Electronics and Mechanics influenced by horizontal technology spillovers?
(3) What is the set of recommendations for government planners and domestic firm managers to take advantage of FDI to improve the labour productivity of domestic firms?
Research scope
This research examines the impact of horizontal technology spillover effects on the labor productivity of domestic firms in both low technology (specifically food processing) and high technology industries (focusing on electronics and mechanics) between 2006 and 2008 Additionally, it explores how controllable factors such as capital intensity, scale, and labor quality influence labor productivity Utilizing the panel least squares method, the study employs various regression models, including pooled regression, fixed effects regression (FEM), and random effects regression (REM), to analyze the data.
Organization of the Research
This research is structured into five chapters: the introduction, a review of the theoretical and empirical framework, a description of the data set along with the research methodology and estimation strategy, a summary and discussion of the model results, and finally, a conclusion that outlines policy implications and identifies research limitations.
CHAPTER 2: LITERATURE REVIEW
Literature Review- The Concepts
Since the early 1980s, Foreign Direct Investment (FDI) has surged dramatically, resulting in a more competitive global market This rise in direct investment flows has significantly contributed to the expansion of international production and the advancement of financial globalization.
Foreign Direct Investment (FDI) plays a crucial role in international economic integration, making it essential to establish a clear definition and global standards for direct investment statistics to effectively analyze FDI activities and their impacts Various organizations, including the OECD (2008), BPM6 (2009), and SNA (2008), offer differing definitions of FDI According to the OECD (2008), which aligns with the economic concepts outlined in BPM6 and SNA, FDI is defined in a comprehensive manner that captures its significance in the global economy.
Foreign direct investment (FDI) signifies a long-term commitment by a resident enterprise in one economy to invest in a business located in a different economy This commitment establishes a lasting relationship between the direct investor and the investment enterprise, allowing the investor to exert significant influence over the management of the enterprise.
Foreign Direct Investment (FDI) is defined by its distinct objectives, which differ significantly from those of portfolio investment Unlike portfolio investors, FDI investors typically seek to have a direct influence on the management and operations of the enterprise in which they invest.
Productivity, as defined by the OECD (2001), is the ratio of output volume to input volume, serving multiple purposes beyond mere measurement It encompasses objectives like tracking technical and efficiency changes, as well as assessing living standards Consequently, various approaches to measuring productivity exist, tailored to specific goals and data availability, including labor productivity, capital productivity, and a combined measure of capital-labor-energy-materials productivity.
Labour productivity is a crucial and easily measurable factor, defined as the ratio of output volume to labour input volume according to OECD (2001) Key influences on labour productivity include capital, intermediate inputs, technology, organizational structure, and changes in efficiency.
In this article, the term "technology" is defined broadly to encompass not only physical equipment but also the knowledge and skills involved in managerial practices and production methods It refers to the various tacit processes through which a firm effectively transforms capital, labor, and materials into a finished product.
Technology spillover from Foreign Direct Investment (FDI) occurs when foreign firms enhance the productivity and efficiency of local non-affiliated businesses in the host country, as noted by Blomstrom and Kokko (1998) Through various FDI projects, foreign investors facilitate the transfer of technology and knowledge, which positively impacts domestic firms As a result, these local businesses experience improvements in productivity and competitiveness.
Horizontal spillovers occur within an industry due to the presence of multinational corporations, facilitated by various mechanisms Firstly, the competition effect compels domestic firms to enhance their management practices and adopt advanced production technologies to remain competitive, as noted by Wang and Blomstrom (1992) Secondly, the demonstration effect, or "learning by watching," allows domestic firms to gain insights from the actions and techniques of foreign companies when new technologies are introduced, as highlighted by Jutta Giinther (2002) Lastly, the employment effect occurs when skilled employees trained in multinational corporation affiliates transition to domestic firms, bringing valuable expertise, as discussed by Fosfuri (1996).
Foreign Direct Investment (FDI) can lead to negative spillover effects that may counteract its positive impacts on domestic firms' productivity When multinational corporations (MNCs) enter a market, they often divert demand from local businesses, potentially leading to a phenomenon known as the market-stealing effect, where domestic firms are forced to reduce production and face declining productivity Additionally, MNCs tend to attract skilled workers away from domestic firms by offering higher salaries and better working conditions, a situation referred to as the skill-stealing effect, as noted by Girma, Greenaway, and Wakelin (2001) This migration of talent further exacerbates the decline in productivity for local firms, highlighting how competitive pressures from MNCs can adversely impact domestic industries.
In summary, foreign firms can have either positive or negative spillover effects on productivity of local competitors depending on a result of balancing out of these offsetting effects.
The Economics Theory
Foreign Direct Investment (FDI) and its spillover effects are crucial topics for scholars and policymakers due to their significant impact on growth in host countries Most theoretical models examining FDI and spillovers operate within a growth theory framework, which highlights the role of FDI in technology transfer The neoclassical theory and endogenous growth theory are recognized as the two primary frameworks that provide in-depth research on FDI spillover effects While these theories present differing assumptions and perspectives, they also offer complementary insights that enhance our understanding of FDI's influence on economic development.
The neoclassical growth model establishes a foundation for understanding long-term economic growth but falls short in addressing a significant source of that growth This limitation is addressed by the endogenous growth theory, which integrates human capital as a key driver of technological advancement, complementing the neoclassical focus on physical capital Rather than viewing these two frameworks as mutually exclusive, analysts can effectively combine elements from both models to gain a comprehensive understanding of the dynamics of long-run economic growth.
The classical theory on growth was developed mainly by Robert M Solow
In 1956, a pivotal working paper titled "A Contribution to the Theory of Economic Growth" emerged, establishing foundational concepts in the study of economic growth The neoclassical theory, based on key assumptions about production functions, asserts that while capital can drive short-term growth, it is technology that plays a crucial role in sustaining long-term growth Additionally, the theory highlights the presence of decreasing returns to each input and the significance of returns to scale in the growth process.
Actually, the production progress augmented by technology is shown as follows:
• Labour is augmented in this research, and the production function appears as:
Y= F (K, AL) (2.2) Dividing both sides of (2.2) by AL, we have the new function as: y K
In which, ~ is output per effective worker and ~ is capital per effective worker
The role of technology in the labor market can be understood through two key assumptions: first, that the employment rate remains constant, leading to a workforce growth rate that matches population growth; and second, that both labor and technology experience growth over time at rates of n and g, respectively Consequently, the growth in effective labor is derived from the sum of these rates, resulting in an overall growth rate of (n + g).
Based on above facts, we get the difference between capital accumulation and depreciation: s~- (J + n +g)(~) (2.4)
In which, s is savings rate and J is depreciation rate
The difference between capital accumulation and depreciation diminishes over time until it reaches zero, indicating that capital per effective worker has attained a steady state By incorporating this steady state level of capital into the production function, we can determine the steady state output per effective worker As the number of effective workers increases at a rate of (n + g), both capital and output also grow at the same rate of (n + g) Thus, technological progress is essential for sustaining long-term growth.
Endogenous growth theory emerged in the 1980s as a critique of exogenous growth theory, which posited that technological advancements are determined externally In contrast, endogenous growth theory asserts that technology is generated within the economy by researchers and innovators While both theories aim to explain long-term economic growth, they differ significantly in the mechanisms they propose to achieve it.
The AK model, a basic form of the endogenous growth theory, builds upon the Solow model with the equation Y = AK, illustrating that output is directly proportional to capital This production function showcases constant returns to the accumulated factor, indicating that a doubling of capital results in a doubling of output.
In line with neoclassical theory, we examine the factors influencing economic growth, focusing on capital depreciation, represented as bK, and capital accumulation, denoted as sY The interplay between these two elements is crucial for understanding the dynamics of economic development.
This equation, together with the Y = AK production function, implies:
L1 YIY = L1KIK = sA - b (2.6) The equation (2.6) implies that output keeps growing as long as sA > b
In this growth model, the long-term growth rate of an economy is determined by the savings rate, whereas in the neoclassical model, the savings rate influences growth only in the short term, with technological progress being the key driver of long-term growth.
In the production function above (Y = AK), the variable K could be interpreted as knowledge (a type of capital) Because new scientific discoveries are
Knowledge production often builds on prior scientific discoveries, which reduces the likelihood of diminishing returns In fact, as Paul Romer (1990) suggested, it may even demonstrate increasing returns.
In conclusion, despite the two fundamental differences between neoclassical and endogenous theories, both share a key principle: technology serves as the primary engine for long-term economic growth.
The Empirical Study
The debate surrounding FDI technology spillovers has yielded mixed empirical results Some studies suggest that FDI can enhance the labor productivity of domestic firms through various mechanisms For instance, the competition effect compels domestic firms to boost productivity to remain competitive (Wang and Blomstrom, 1992), while the demonstration effect leads to the adoption of new technologies introduced by multinational corporations (MNCs), resulting in improved production (Jutta Gunther, 2002) Additionally, the employment effect occurs when skilled employees transition from MNCs to domestic firms (Fosfuri, 1996) Research by Caves (1974) on 23 Australian manufacturing industries indicates that a higher share of foreign subsidiary employment correlates with increased productivity in local firms Similarly, Globerman (1979) found positive impacts of foreign shares on labor productivity in Canadian manufacturing industries Other studies supporting positive FDI spillover effects include Blomstrom and Persson (1983) in Mexico and Blomstrom and Sjoholm (1998) in Indonesia.
Some studies indicate that technology spillovers can have negative effects on the labor productivity of domestic firms This decline may be attributed to two main factors: the market stealing effect, where multinational corporations (MNCs) divert demand away from local businesses, and the skill stealing effect, which undermines the competitive advantage of domestic firms.
Research by Haddad and Harrison (1993) on Moroccan manufacturing from 1985 to 1989 reveals no significant positive correlation between increased productivity in domestic firms and foreign presence in the sector Additionally, studies by Aitken and Harrison (1999) in Venezuela and Konings (2001) in Bulgaria report negative or negligible spillover effects, highlighting the potential skill stealing effect where domestic firms experience productivity declines due to the loss of their best workers to foreign competitors.
Contradictory findings regarding FDI spillover effects can be attributed to factors such as study design, data characteristics, and the measures used to assess foreign presence A significant factor is the technology gap between multinational corporations (MNCs) and domestic firms; a larger technology gap typically results in diminished spillover benefits for domestic firms (Tamotsu Nakamura, 2002; Klaus E Meyer, 2004) Consequently, spillover effects in developing countries often differ from those observed in industrialized nations Research indicates that while some studies in developing countries, such as Haddad and Harrison (1993) for Morocco and Aitken and Harrison (1999) for Venezuela, report null or negative spillover effects, studies in more advanced economies, like Haskel et al (2002) for the United Kingdom and Keller and Yeaple (2003) for the United States, tend to find positive spillover effects.
In developing countries like Vietnam, it is essential to assess whether evidence supports the presence of productivity spillovers Research by Rossitza B Wooster and David S Diebel (2010) indicates that empirical studies on Foreign Direct Investment (FDI) spillovers in these nations generally estimate intra-industry productivity spillover effects using a standard equation format.
Productivit y Measure = 00 + L ojxj + okForeign Presence + £ ( ) j=l 2.7 r
In equation (2.7), the key parameter is the estimated coefficient on the measure of foreign presence, t5 - k, which reflects the impact of foreign presence on the productivity of domestic firms within an industry.
This meta-analysis examines three types of productivity measures: sector output, total factor productivity, and labor productivity, with the choice of measure influenced by data availability and reliability Aitken and Harrison (1999) utilize sector output, while Chuang and Lin (1999) opt for total factor productivity, and Blomstrom and Sjoholm (1998) focus on labor productivity.
Various studies utilize distinct methods to assess foreign presence in the market, including sector output share as demonstrated by Blomstrom and Sjoholm (1998), sector capital share analyzed by Chuang and Lin (1999), and sector employment share examined by Aitken and Harrison.
Empirical studies reveal varying definitions of foreign-invested firms and the types of data utilized Some researchers classify any firm with foreign equity exceeding zero as foreign-invested (Konings, Jozef, 1999; Damijan et al., 2003), while others, like Djankov and Hoekman (2000), set the threshold at 20% foreign equity Additionally, certain Chinese studies adhere to the government's definition of foreign firms, as noted by Buckley et al (2002) and Xiaming Liu et al (2001).
Referring to types of data employed, some old studies tend to use cross- section data aggregated at industry level (Blomstrom and Persson, 1983; Kokko,
Recent studies have increasingly utilized firm-level panel data, as seen in the works of Sinani and Meyer (2004) and Damijan et al (2003), contrasting with earlier research that focused on cross-sectional data The choice between using cross-sectional or panel data, as well as the level of aggregation—whether at the industry or firm level—plays a crucial role in the analysis.
• on the data availability and reliability; however, panel data with aggregation at firm level are used commonly in rencent studies
Empirical studies on foreign direct investment (FDI) spillover effects reveal mixed results Positive spillover effects have been reported in Indonesia by Blomstrom and Persson (1999), in China by Buckley et al (2002), and in Vietnam by Thuy (2005) Conversely, other research indicates no evidence or even negative spillover effects, as shown by Aitken and Harrison (1999) in Venezuela, Kathuria (2002) in India, Sadik and Bolbol (2003) in Arab countries, and Sinani and Meyer (2004) in Estonia.
A meta-analysis conducted by Rossitza B Wooster and David S Diebel in 2010, which examined 141 regression results from 32 studies, reveals weak evidence supporting the notion that domestic firms in developing countries benefit from foreign direct investment (FDI) within the same sector This suggests that intrasectoral spillovers from FDI in these regions may be largely absent.
In Vietnam, research on the spillover effects of Foreign Direct Investment (FDI) remains limited A notable study by Le Thanh Thuy (2007) examines the impact of FDI on the productivity of domestic firms across various industries during two periods: 1995-1999 and 2000-2002 Utilizing industry-level panel data from 29 sectors, including four in mining and quarrying, 23 in manufacturing, and two in electricity, gas, and water supply, the study reveals significant positive spillover effects in the earlier period, while the later period shows insignificant spillovers.
In 2002, Vietnam underwent significant structural reforms that led to substantial positive effects from foreign direct investment (FDI), characterized by notable demonstration and competition impacts However, as the market stabilized in subsequent years, the outcomes shifted, reflecting a market stealing effect during this later period.
In a different way, Nguyen Thi Tue Anh et al (2006) uses cross section- firm level data to examine the impacts of FDI spillover on labour productivity of
The study analyzed the impact of foreign direct investment (FDI) on manufacturing firms across various sectors, including food processing, textiles and footwear, and mechanics and electronics It identified four key channels of spillover effects: labor turnover, technology diffusion and transfer, production linkages, and competition Utilizing a quantitative model, the research revealed that positive spillover effects from FDI are generally present in these industries.
Analysis Framework
The neoclassical perspective on spillover effects highlights the indirect impacts of foreign direct investment (FDI) on the productivity of domestic firms, emphasizing the significance of externalities Numerous studies, including those by Kokko et al (1996), Xiaming et al (2004), Javorcik (2004), and Wei and Liu (2001), have utilized the neoclassical growth framework to explore these productivity enhancements stemming from FDI.
According to Xiaming Liu et al (2001), the "productivity spillover model" illustrates that local firms' labor productivity (lp) is affected by several factors including capital intensity (ci), labor quality (lq), firm size (fs), foreign presence (ifp), and additional explanatory variables (ov) like industry concentration, R&D intensity, and the technology gap between local and foreign firms.
According to Romer (1994), knowledge is a crucial factor of production, alongside traditional inputs like labor and capital Within the framework of an endogenous growth model at the firm level, firm-specific capital serves as the driving force behind productivity growth This concept has been explored in the works of Lucas (1988), Romer (1990), and Ehrlich et al (1994), highlighting the significance of knowledge in enhancing firm performance Recently, Z Liu has further contributed to this discussion.
(2008) specifies the firm's production function as:
In a joint venture, A1 denotes exogenous common technical factors, while B1 refers to the productivity parameter associated with advanced technology introduced by the foreign partner Additionally, L1 and K1 represent labor and capital inputs, respectively, and H1 signifies the stock of firm-specific capital Lastly, Mr indicates the proportion of managerial time allocated to the venture.
The endogenous growth framework provides insights into technology spillovers at the firm level; however, the neoclassical model remains a highly suitable approach for examining the impact of foreign direct investment (FDI) on labor productivity.
This study employs a neoclassical production function to analyze how the labor productivity of domestic firms is influenced by factors such as capital intensity, labor quality, economies of scale, foreign presence in the industry, and time period Utilizing firm-level panel data, the research investigates the horizontal technology spillover effects of foreign direct investment (FDI) on labor productivity in both low technology industries, like food processing, and high technology sectors, such as electronics and mechanics.
CHAPTER 3: RESEARCH METHODOLOGY
Description ofVariables
3.2.1 Dependent variable: Labour productivity (LABPRO)
In estimating horizontal productivity spillover effects, various productivity measures can be utilized, including sector output, labor productivity, or total factor productivity This research specifically defines productivity in terms of labor productivity, calculated as value added (VA) per labor Value added is determined by subtracting intermediate costs (IC) from gross output (GO) for the year, with IC derived by multiplying GO by the industry's IC ratio This ratio is sourced from the General Enterprise's Cost Survey (GECS), conducted every five years by the General Statistics Office (GSO) of Vietnam.
According to GSO, labour productivity is calculated by dividing value added by the average number of labour in firms and its unit value is million VND
3.2.2.1 Principal independent variable: Foreign presence (FDISPILL)
In the literature, foreign presence in firms is typically assessed through three key metrics: the share of total sector employment, total sector capital, and total sector output This research specifically focuses on measuring foreign presence by examining its share of total sector output.
Blomstrom and Sjoholm (1998) define a proxy for foreign direct investment (FDI) presence as the percentage of total gross output produced by foreign-owned establishments within a five-digit industry This no-unit variable is collected annually at the end of each year.
The literature review reveals a complex relationship between labor productivity and foreign presence, influenced by study design, data characteristics, and measurement choices This research posits that foreign presence negatively impacts labor productivity in low-technology industries, such as food processing, while having no significant effect on labor productivity in high-technology sectors like electronics and mechanics.
3.2.2.2 Controllable independent variables 3.2.2.2.1 Capital intensity (CAPIN)
Capital intensity plays a crucial role in determining a firm's productivity, as it reflects the amount of physical capital available per employee Research by Blomstrom and Sjoholm (1998) and Xiaming Liu et al (2001) suggests that higher capital intensity is likely to enhance labor productivity within firms.
This research defines capital intensity as the ratio of average fixed visible capital assets to the average number of employees within a firm, measured in millions of VND and sourced directly from the GSO at year-end It is posited that there is a positive relationship between capital intensity and labor productivity.
Economies of scale play a crucial role in influencing firm performance, as they reflect a company's market power within its industry Blomstrom and Sjoholm (1998) define economies of scale as the relationship between a firm's size and its industry, measured by the firm's revenue relative to the average revenues of its five-digit industry classification A firm that generates higher revenue compared to the industry average is presumed to have a competitive advantage.
- economies of scale and thus, higher productivity In other words, economics of scale is positively related to labour productivity
Labour quality is defined by various authors as the ratio of skilled professionals—such as managers, scientists, engineers, and technicians—to the total workforce, or the comparison of white-collar to blue-collar workers (Kokk:o, 1996; Blomstrom et al., 1998; Liu and Wei, 2001; Li et al., 2004) Additionally, Nguyen Thi Tue Anh et al (2006) describe labour quality as the proportion of employees who have completed at least college or vocational training within a firm This variable is anticipated to have a positive coefficient, indicating its beneficial impact on the labour productivity of firms.
This research measures labor quality by calculating the ratio of employees who have completed at least a college education or vocational training to the total workforce within firms This non-unit variable is collected annually at the end of each year.
Population distributions can vary across different time periods, necessitating the use of year dummy variables to account for these temporal effects in research This study analyzes data from three years, specifically 2006 to 2008, with the year 2006 serving as the base year to control for changes that are consistent across firms but fluctuate over time.
Model Specification
This study explores the connection between the labor productivity of domestic firms and horizontal technology spillovers from foreign direct investment (FDI) by utilizing a neoclassical production function, following methodologies established in prior research (Blomstrom and Sjoholm, 1998; Xiaming Liu et al., 2001).
LABPRO = j(CAPIN, SCALE, SKILL, FDISPILL)
In the logarithm form, the model can be expressed as follows: ln(LABPROijt) = !t! [95% conf lncAPIN • 7122745 0207361 34.35 0.000 6716117 SCALE: 3224544 015154 21.28 0.000 2927379 SKILL -.565974 1146379 -4.94 0.000 -.7907751 FDISPILL • 5364627 0985264 5.44 0.000 3432558 _cons 01.98137 0869001 0.23 0.820 -.1505945
Random effects model result in Food Processing industry
lttreg 1~ lrcAPIN SCALE SKILL FDISPILL, re
Random-effeCLS GLS regression Group variable: i
Random effects u_i - Gaussian corr(u_i, X) = 0 (assumed) lnLABPRO coef Std Err lncAPIN 6451168 021772 SCALE 2992742 0177837 SKILL -.1838828 1359229 FDISPILL 3780614 1492781 _cons 2923689 0987309 sigma_u 90557632 sigma._e 48763046 rho • 77522046 (fraction of z
Number of obs Number of groups obs per group: min avg max wald chi2(4) Prob > chi2
2.3 Pool or Random: Breusch-Pagan Lagrange multiplier test
Breusch-Pagan Lagrange multiplier test
Breusch and Pagan Lagrangian multiplier test for random effects lnLABPRO[i,t] = Xb + u[i] + e[i,t]
Test: lnLABPRO e u var(u) 0 var sd = sqrt(var)
The LM test results indicate a p-value less than 0.05, leading to the rejection of the null hypothesis Consequently, random effects regression is deemed more suitable than pooled OLS regression for the analysis.
Fixed effects model result in Food Processing industry
Jttreg lrLABPRO lrx:APIN SCALE SKILL FDISPILL, fe
Fixed-effects (within) regression Group variable: i
R-sq: within = 0.2729 between= 0.3965 overa1l 0.3845 corr(u_i, xb) 0.0099
Number of obs Number of groups obs per group: min
148.05 0.0000 lnLABPRO coef std Err P>lt! [95% conf Interval] lncAPIN SCALE SKILL FDISPILL _cons
• 5107479 1996056 -.14352 -2.979142 6368658 sigma_u sigma_e rho 83759375 (fraction of variance due to u_i)
2.5 Fixed or Random: Hausman test
0174889 0185031 1353828 5734603 b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(4) Prob>chi2
As the result of Hausman test in the table above, the p-value is smaller than 0.05; that means fixed effects model is more appropriate than random effects model
2.6 Including time effects in a fixed effects model
Fixed effects model result with time fixed effects in Food Processing industry
• xi: xt:reg lrL.ABPRO lrcAPIN SCALE SKILL FOISPILL i t: • fe i t: _It_l-3 (naturally coded; _It_l omitted)
R-sq: within = 0.2814 between = 0.4456 overall 0.4294 corr(u_i, xb) = 0.0874 lnLABPRO coef lncAPIN • 5521868 SCALE 2532728 SKILL 3584056 FDISPILL -1.366945
_It_2 1091008 _It_3 0584056 _cons • 8601855 sigma u 1.0656822 sigma e 48506139 rho 82838009
Number of obs Number of groups obs per group: min = max avg
As the result above, p-value is smaller than 0.05 and this reject the null hypothesis; that means time fixed effects must be included a fixed effects regression
Modified wald "tes1: for groupwise he1:eroskedas"tici1:y in fixed effect regression model
HO: signa(i)A2 = si!JII