Essays on financial activities and industrial change the study case in vietnam

GENERAL INTRODUCTION

Background of the study

The Agenda for Development (A/RES/51/240) defines development as a process aimed at enhancing the quality of life for all individuals Key elements for achieving sustainable development include economic, social, and environmental growth Furthermore, the Sustainable Development Goals (SDGs) emphasize the need to empower domestic financial institutions to broaden access to banking, insurance, and financial services for everyone This study specifically examines sustainable economic development in Vietnam.

Rural development is crucial for alleviating poverty in Vietnam, where the government has implemented various programs, including microcredit and agricultural extension services, to support households Since the mid-1980s, Vietnam has successfully transitioned from one of the poorest countries to a middle-income nation (Banker & Ungor, 2019) Recently, the country has transformed its economic structure through industrialization and modernization, leveraging the comparative advantages of its industries and services.

Objectives

This study aims to achieve two main objectives:

Between 2000 and 2015, Vietnam underwent significant economic structural changes, highlighting key sectors that drove growth during this period The analysis focuses on identifying these pivotal sectors and understanding the dynamics of Vietnam's evolving economy.

This article explores the financial activities of small and medium-sized enterprises (SMEs) and rural households in Vietnam It aims to identify the relationship between corporate finance, specifically financial leverage, and investment decisions within Vietnamese SMEs Additionally, it evaluates how membership in farmer unions influences the financial dynamics of rural households.

Contribution of the study

This study enhances the current understanding of rural development in Vietnam by examining the role of social organizations in these areas Furthermore, it adds to the body of knowledge on the interplay between corporate finance and investment decisions within emerging markets.

This study uses several methodologies to address the problem such as causal inferences, input-output decomposition analysis which have received much attention in the current research literature.

Structure of the study

This study is structured into four chapters, beginning with an overview of Vietnam's economic development context Chapter 2 examines the significance of the food supply chain in Vietnam from 2000 to 2015 through a decomposition Input-output approach Chapter 3 analyzes the relationship between corporate finance and the investment decisions of Small and Medium Enterprises (SMEs) in Vietnam, utilizing various econometric techniques, including Logit, Tobit, and Fractional logit models Finally, Chapter 4 provides an empirical estimation of how farmer union membership affects household production and credit volume, employing the Propensity Score Matching approach.

HOW IMPORTANT A FOOD SUPPLY CHAIN IN VIETNAM

Introduction

Many economists, such as Kuznets (1979), Lin and Monga (2011), Uy et al

Economic restructuring plays a crucial role in a country's development and growth, as highlighted by Vu and others in their research In Vietnam, recent years have seen significant transformation through industrialization and modernization, leveraging comparative advantages in various industries and services After three decades of renovation, Vietnam has achieved remarkable progress, successfully transitioning from a low-income to a middle-income country, as noted by Barker and Ungor.

A substantial body of research has explored changes in economic structure, including notable studies by Leotief (1941), Feldman et al (1987), and others Many of these studies adopt a macroeconomic perspective to examine aggregate indicators like employment, growth rates, and GDP (Hacks, 1989; Skolka, 1989; Schoonbeek, 1990; Pham Quang Ngoc & Mohnen, 2004; Rahmaddi & Ichihashi, 2013) For instance, Pham Quang Ngoc & Mohnen (2004) utilized a multisectoral model to analyze the relationship between Vietnam’s economic growth and structural change, referencing input-output tables from 1989, 1996, and 2000 Additionally, K.M V (2017) highlighted the significant positive impact of structural change on GDP growth While traditional economic analyses often overlook the industrial approach, it remains a crucial tool for understanding structural change This study aims to leverage the industrial approach to identify key sectors and examine the evolution of Vietnam's economic structure through its input-output tables from 2000 onwards.

2015 Backward-forward linkages and the decomposition approach are applied to

4 analyze the IO tables The major findings are used to explain the role of certain subsectors and reveal the trends in the economic structure of Vietnam’s economy

This paper significantly contributes to the literature on structural economic change by analyzing the Vietnamese economy through industrial analysis, decomposing output changes into technology and final demand for each sector This approach allows for an investigation into the key factors influencing total input changes Additionally, it is the first study to utilize four Vietnamese input-output data tables to compare the economy's structure across three distinct periods, providing insights into Vietnam's structural economic changes over a 15-year span from 2000 to 2015.

Vietnam has undergone significant structural economic changes, as reported by the General Statistical Office (GSO) in 2016 The country has decreased its reliance on agriculture, with its contribution to GDP falling from 23.24% in 2000 to 17% in 2015, a decrease of 6.2% Additionally, the industrial and construction sector's share dropped by 4.88%, from 38.13% to 33.25% Conversely, the services sector has seen growth, increasing its share of GDP by 1.1%, from 38.63% to 39.73%.

Figure 1 illustrates the share of four main sectors in Vietnam’s GDP from

Between 1986 and 2009, significant shifts occurred in various economic sectors, particularly agriculture and manufacturing The agricultural sector's contribution to GDP plummeted from 34% in 1986 to 17% in 2009, while the manufacturing sector saw growth from 17% to approximately 25% The services sector emerged as the dominant force, increasing its share from over 46% in 1986 to 54% in 2009 In contrast, the mining and quarrying sectors consistently contributed less than 6% to GDP during this period.

Figure 1: Shares of GDP by Sectors, 1986–2009

The food supply chain is crucial for rural development, providing essential materials for various industries (Marsden et al., 2000; Kastrinaki & Stoneman, 2011) It encompasses all processes from production to distribution of food and drink (Aramyan & Van Gogh, 2014) In Vietnam, the agricultural sector remains the primary supplier of raw materials to other industries, particularly the food and textile sectors, with the food sector heavily relying on agricultural inputs (Dieu TTM, 2006).

Input-Output (IO) analysis illustrates the interconnectedness of industries, households, and government within an economy Numerous studies, including those by Skolka (1989), Franke and Kalmbach (2005), and Marconi et al (2016), have utilized IO analysis to explore national economies Additionally, research by Hongsakhone and Ichihashi (2019) has focused on the interdependencies among households In Vietnam, several investigations, such as those conducted by T Bui et al., have examined shifts in the economic structure using IO tables.

In their research, T Bui and Kobayashi (2012) explored interindustry linkages between manufacturing and nonmanufacturing sectors using three Input-Output tables from 1989, 1996, and 2000 Their findings indicated that the manufacturing sector exhibited stronger internal linkages compared to nonmanufacturing sectors Subsequent studies by Phong N.V (2013), Nguyen P Thao (2014), and Ha, N.H.P & Trinh, B (2018) further contributed to this discourse, highlighting the significance of these linkages in understanding economic dynamics.

Current research focuses on the IO decomposition method; however, there is no study that utilizes four I-O tables Most existing literature compares economic data over long intervals, typically five years, which may not accurately reflect changes in economic structure since such structures do not significantly evolve in short periods.

Our research highlights significant output growth in the agriculture, hunting, forestry, fishing, food and beverage, and tobacco product sectors due to shifts in final consumption demand Additionally, the machinery and equipment sector saw notable increases driven by changes in investment and export demands, along with technological advancements in intermediate inputs Conversely, the wholesale and retail trade and repair sector faced a dramatic decline in final demand for consumption, investment, and exports over the past five years (2010-2015).

This article is structured to provide a comprehensive overview of the research on Vietnam's economy, beginning with an introduction to the research statement and relevant background in Section 1 Methodology is detailed in Section 2, followed by an explanation of the data utilized in Section 3 The empirical results are discussed in Section 4, while Section 5 offers a discussion on the findings Finally, Section 6 concludes with key remarks.

Methods and Models

To identify the key sectors in the Vietnamese economy and investigate its structural transformation over the last 15 years, we analyze Vietnam’s IO tables for

2000, 2005, 2010 and 2015 We aggregated the 2015 data into 34 industries to ensure that the IO tables for all years were comparable

Where X is total output; [𝐼 − (𝐼 − 𝑀)̂ 𝐴] −1 is the Leontief inverse matrix;

A is the input coefficient matrix; I is the 34x34 identity matrix; 𝑀̂ is a 34 x 34 diagonal matrix with diagonal elements 𝑚 𝑖

𝑀)̂ is self-sufficient rate matrix; 𝐹 𝑑 is vector of domestic final demand; 𝐸𝑋 is vector of total export

Backward and forward linkages, concepts introduced by Rasmussen in 1956, are essential for identifying key sectors within an economy Backward linkage refers to the connections a sector has with others through its input purchases, illustrating how a one-unit increase in sector i affects other sectors Conversely, forward linkage focuses on the relationships a sector maintains with others through its output sales Both linkages are quantitatively assessed using specific equations, as outlined by Miller and Blair in 1985.

Where: bij are the inverse matrix elements [𝐼 − (𝐼 − 𝑀)𝐴̂] −1 , n is the number of sector (n4)

According to the theory of IO model, the equation of output equals the matrix product of the Leontief inverse (B) and the vector of final demand (F) can be expressed as:

The change of output can be expressed in the following ways:

⏟ (7) Technology change Final demand change

Where: 𝑩 = [𝑰 − (𝑰 − 𝑴 ^ )𝑨] −𝟏 is the Leontief inverse matrix, F is the vectors of final demand

This study utilizes the structural decomposition method established by Dietzenbacher & Los (1998), where the total output change is analyzed as the sum of changes in final demand and technology Subsequently, the research further breaks down the changes in final demand by examining its components, including consumption (C), investment (I), and exports (EX).

Data

This study utilized Vietnam's Input-Output (IO) tables from 2000, 2005, 2010, and 2015, as published by the Organization for Economic Co-operation and Development (OECD) The IO tables detail the transactions of goods and services across 34 sectors in current prices (USD million) for the years 2000, 2005, and 2010, while the 2015 table includes 36 sectors To maintain consistency in our analysis, the 2015 IO table was adjusted to align with the 34-sector format Table 1 below lists the 34 sectors included in this research.

Table 1: Sectors selected for the study

1 Agriculture, hunting, forestry and fishing

3 Food products, beverages and tobacco

4 Textiles, textile products, leather and footwear

5 Wood and products of wood and cork

6 Pulp, paper, paper products, printing and publishing

7 Coke, refined petroleum products and nuclear fuel

10 Other non-metallic mineral products

14 Computer, electronic and optical equipment

16 Motor vehicles, trailers and semi-trailers

19 Electricity, gas and water supply

21 Wholesale and retail trade; repairs

27 Renting of machinery and equipment

30 Public administration and defense; compulsory social security

33 Other community, social and personal services

34 Private households with employed persons

Results

Table 2 displays the output structure based on the 34 sectors The total output of the Vietnamese economy $66,545.9 million in 2000 and $570,059.6 million in

Between 2000 and 2015, the agricultural sector's share of the economy decreased from 21.15% to 13.07%, while the wholesale, retail, trade, and repairs sector also saw a decline from 11.56% to 5.02% In contrast, the food products, beverage, and tobacco sector experienced growth, rising from 11.21% to 12.42% during the same period, marking a significant increase of 3.72%.

10 share of the textiles, textile products, leather, and footwear sector to Vietnam GDP over the 15-year period 2000-2015

Table 2: Total output by sectors (unit: millions of US dollars)

Agriculture, hunting, forestry and fishing

Food products, beverages and tobacco

Textiles, textile products, leather and footwear

Wholesale and retail trade; repairs 7690.2 11.56 13873 9.87 30033.9 10.48 28630.4 5.02

Coke, refined petroleum products and nuclear fuel

Renting of machinery and equipment

Electricity, gas and water supply 1363.5 2.05 2761.7 1.97 5673.5 1.98 12531.3 2.20

Motor vehicles, trailers and semi- trailers

Other non-metallic mineral products 1351.1 2.03 3836.3 2.73 7036 2.45 9715.3 1.70

Pulp, paper, paper products, printing and publishing

Wood and products of wood and cork 297.5 0.45 1382.6 0.98 2258.2 0.79 6791.4 1.19

Other community, social and personal services

Private households with employed persons

Table 3 presents the backward and forward linkage results for 34 industries across the years 2000, 2005, 2010, and 2015, as derived from equations (2) and (3) These linkages assess the intersectoral connections of each industry with others Backward linkage reflects the influence of a specific industry on others, while forward linkage reveals how an industry is affected by external sectors, as outlined by Chenery and Watanabe (1958).

According to Table 3, the agriculture, hunting, forestry, and fishing sector is the leading sector in the Vietnamese economy, as both its backward and forward

The analysis reveals that all 13 linkages exceed 1 across all periods, indicating robust interconnections among agriculture, hunting, forestry, and fishing with other sectors in terms of both input demand and output supply Additionally, the pulp and paper industry remains a significant sector throughout the years Notably, the food products, beverages, and tobacco sector underwent considerable transformation over the 15-year span; in 2000, it heavily relied on interindustry supply, but by 2005, it emerged as a vital component of the Vietnamese economy.

BL FL BL FL BL FL BL FL

Wood and products of wood and cork

Pulp, paper, paper products, printing and publishing

Other non-metallic mineral products

Motor vehicles, trailers and semi- trailers

Electricity, gas and water supply

Renting of machinery and equipment

Wholesale and retail trade; repairs

Public administration and defence; compulsory social security

Private households with employed persons

Using equation (7), we analyzed the total output change by breaking it down into the impacts of technology and final demand across three distinct periods: 2000-2005, 2005-2010, and 2010-2015 The findings presented in Table 4 indicate that the agriculture, hunting, forestry, fishing, and food, beverage, and tobacco product sectors were significantly affected during these years.

From 2000 to 2015, the machinery and equipment sector saw significant technological advancements, leading to considerable output growth However, during the last period from 2010 to 2015, the wholesale and retail trade and repair sector faced a sharp decline in total output.

Between 2000 and 2005, the food sector experienced the most significant technological advancements, followed by the mining and quarrying sector from 2005 to 2010, and the textiles sector from 2010 to 2015 Notably, the food sector maintained the second-highest level of technological change in the latter period Additionally, the agriculture sector demonstrated the most substantial change in final demand across all three periods, with the food, wholesale and retail, textiles, and construction sectors following closely behind.

2015), there was a drastic decline in the wholesale and retail sector’s final demand

In contrast, the machinery sector presented a rapid increase in its total output and final demand

Using equation (8), we analyzed the changes in final demand, breaking them down into consumption, investment, and exports, as detailed in Table 5 The agriculture sector exhibited the most significant consumption changes across all three periods, followed by the food and wholesale sectors In contrast, the textiles sector experienced the highest export changes during the same timeframe, with the agriculture and food sectors also showing strong export performance Notably, the agricultural sector ranked third in export changes from 2000-2005 and 2015-2010, and second from 2005-2010, while the food sector ranked fifth, fourth, and second in export changes during 2000-2005, 2005-2010, and 2010-2015, respectively Additionally, construction demonstrated the largest investment change over the 15-year span, and the machinery sector's growth in the last five years was significantly influenced by changes in final investment and export demand.

Some main findings are clear from the tables First, the agricultural sector and food sector played important roles throughout the 15 years considered The

The significant changes in total output within the textile and machinery sectors of Vietnam's economy are primarily attributed to shifts in final consumption demand The textile sector has emerged as a key player, experiencing growth driven by increased exports In contrast, the machinery sector saw a notable rise in total output from 2010 to 2015, influenced by changes in final investment, export demand, and advancements in technology However, the wholesale and retail sector faced a marked decline in final investment and export consumption demand during the same five-year period.

Table 4: Top 5 sectors with the greatest change in total output in terms of changes in technology and final demand in the period 2000-2015 (unit: millions of US dollars)

Food products, beverages and tobacco 40716.8

Textiles, textile products, leather and footwear 37785

Agriculture, hunting, forestry and fishing 30579.7

Textiles, textile products, leather and footwear 4959.9

5379.91 Food products, beverages and tobacco

Coke, refined petroleum products and nuclear fuel 7475.96

Renting of machinery and equipment 6200.10

Wholesale and retail trade; repairs 16423.7

Table 5: Top 5 sectors with the greatest change in final demand in terms of the change in consumption, investment and exports in the period 2000-2015

(unit: millions of US dollars)

Motor vehicles, trailers and semi-trailers 1094.89

Other non- metallic mineral products 928.99

Computer, Electronic and optical equipment 4784.90

Motor vehicles, trailers and semi-trailers

Discussion

The agriculture sector has remained pivotal in Vietnam's economy from 2000 to 2015, significantly transforming since the Doi Moi policy reforms initiated in the 1980s The shift from collective farming to household-based farming, established by Decree No 10 in April 1988, allowed farmers greater autonomy over their production While agricultural cooperatives improved rural infrastructure, they failed to ensure equitable income distribution among members The 1990s saw further reforms that reduced government control over agricultural production, including the easing of input trading restrictions and the promotion of free domestic rice trade These changes enabled farmers to access production inputs more easily and sell their products freely, leading to notable growth in both consumption and exports in the agriculture sector between 2000 and 2005.

Under Decree 5 from the 4th Party Congress, the Vietnamese government aims to shift the agricultural economic structure towards large-scale commodity production that integrates with the processing industry and market demands The agricultural sector plays a crucial role in Vietnam's food supply chain, supplying essential inputs for food processing enterprises Analysis of four Input-Output tables reveals that the agriculture sector primarily sells its products to the food products, beverages, and tobacco sectors as of 2015.

61%, 2010 – 46%, 2005 – 46%, 2000 – 38%) Additionally, the food sector mostly purchases inputs from the agriculture sector (approximately 50%)

Vietnam's textile sector has experienced rapid growth, significantly influenced by free trade agreements Between 2000 and 2015, it emerged as one of the fastest-growing industries globally, achieving an impressive annual growth rate of around 6%, positioning Vietnam as a leading exporter in the textile market.

Vietnam has emerged as the second-largest supplier of textiles and garments to major markets including the United States, the European Union, and Japan (Pertiwi & Sukmawani, 2017) Historically, prior to its independence in 1975, the Vietnamese government fully controlled the textile and garment industries, primarily exporting products to the Soviet Union (A.N Tran, 1996) Post-independence, the industry shifted towards self-management, and the signing of a bilateral free trade agreement with the US in 2000 marked a significant increase in textile and garment exports globally Vietnam's accession to the WTO in 2007 further opened markets in the US, EU, and China, providing substantial benefits to the industry (CIEM, 2010) Additionally, Vietnam has entered several trade agreements with Australia, South Korea, and Japan within the ASEAN framework, enhancing its textile export capabilities Recent trade agreements, including the EU-Vietnam Free Trade Agreement (EUVFTA) and the Trans-Pacific Partnership (TTP), are expected to further bolster the growth of Vietnam's textile sector.

In Vietnam, state-owned corporations maintain significant control over key industries, such as the Vietnam Coal and Minerals Industries Corporation (Vinacomin) in mining, the Vietnam Tobacco Corporation (VINATABA) in tobacco, and the Vietnam Food Association (Vinafood 1 and Vinafood 2) in rice products However, some textile enterprises, like Vinatex and Garco10, are increasingly becoming independent from government influence Despite these shifts, challenges persist in the overall development of various sectors.

23 of particular sectors under free trade Table 6 below provides some examples of the large companies that belong to the aforementioned sectors in the Vietnamese market

Our study aligns with previous research by Dang et al (2019) and Ha and Trinh (2018), which analyzed Vietnam's economic structure using 2012 and 2016 input-output tables Their findings highlighted the significant influence of the agriculture, food, oil and gas, and manufacturing sectors on the input demands of other industries Despite structural shifts in the economy from 2000 to 2015, agriculture and food sectors remained vital However, challenges persist, as economies dominated by primary industries like agriculture may face vulnerabilities due to decreasing returns to scale To foster economic growth, innovation and new technologies are essential; without them, expanding production becomes challenging Consequently, primary sectors may hinder progress, keeping countries in poverty (Chang, 2003; Reinert, 2007) Looking ahead, industries like machinery and textiles could emerge as leading sectors in Vietnam's economy.

Table 6: Some representatives large enterprises in the selected industries in Vietnam Table 6: Some representative large enterprises in selected industries in Vietnam

Textile industry Garment Industry Footwear

Vietnam National Textile and Garment Group (VINATEX)

Pou Yuen Vietnam Limited Liability Company

Duc Quan Investment and Development Joint Stock Company Viet Tien Garment Corporation

TaeKwan Vina Industrial Limited Liability Company

Dam San Joint Stock Company Garco 10 Corporation

Hwaseung Vina Limited Liability Company

Phu Bai Spinning Mill Joint Stock

Chang Shin Viet Nam Limited Liability Company

Hoa Tho Textile Garment Joint Stock Company

Pou Sung Vietnam Limited Liability Company

Saigon Beer-Alcohol-Beverage Corporation (Sabeco)

Acecook Viet Nam Joint Stock Company

Vietnam Tobacco Corporation (Vinataba) Saigon Tobacco Limited Company

Heineken Vietnam Vinh Hoan Joint Stock Company Vinataba Thang Long Hanoi Beer, Alcohol and Beverage

Bien Dong Seafood Limited Liability Company

Hanoi Liquor Joint Stock Company Minh Phu Seafood Corporation

Binh Tay Wine Joint Stock Company

VIETNAM AGRIBUSINESS Limited liability Company

Rice products Fishery and aquaculture Wood

Southern Food Corporation – VINAFOOD II Minh Phu Corporation – MPC Hoa Net Limited liability company Northern Food Corporation –

VINAFOOD I Vinh Hoan Corporation– VHC Nitori Furniture Vietnam Tân Thạnh An Limited Liability

Bien Dong Seafood Limited Liability Company An Cuong Woodworking materials Kien Giang Import and Export Joint

Shing Mark Vina Limited liability Company

Tan Dong Tien Joint Stock Company Hung Vuong Corporation – HVG DONGWHA Corporation

Hoa Phat Joint Stock Company Thaco Group Kubota Tractor Corporation Gang Thép Thái Nguyên Joint Stock

Company Toyota CLAAS KGaA GmbH

VISCO Joint Stock Company Honda

Vietnam Engine and Agricultural Machinery Corporation (VEAM)

Dana Joint Stock Company Ford Thaco Corporation

Viet Duc Joint Stock Company GM Vietnam Truong Hai Auto Corporation (THACO)

Vinacomin Masan resources Nui Phao Thach Khe (Vinacomin)

The Vietnamese agricultural machinery market is projected to grow from USD 425 million in 2025 to USD 560 million by 2030, with a CAGR of 11.5% This growth is driven by the increasing demand for agricultural productivity and the adoption of modern farming equipment to enhance efficiency and reduce labor costs Labor shortages, particularly in rural areas, have led to a rise in mechanization, such as the use of combine harvesters, which significantly improve harvesting efficiency The market is characterized by a diverse range of machinery, including tractors and harvesting equipment, tailored to meet the needs of small- to medium-sized farms Key players in the market, including Kubota and CLAAS, are focusing on innovations and strategic partnerships to expand their presence As Vietnam's agricultural sector faces demographic shifts and rising labor costs, the reliance on modern machinery is becoming essential for maintaining competitiveness and meeting growing domestic and export demands.

Conclusions

Identifying leading industries is crucial for poverty reduction and economic growth in developing countries For instance, the IT sector in the U.S., the automobile industry in Japan, and the electronics sector in South Korea have significantly contributed to their economies Despite Vietnam's rapid growth over the past 15 years, it is still classified as a developing nation This paper aims to pinpoint the primary industries driving Vietnam's economic development by utilizing the decomposition method on a series of input-output tables from 2000 to 2015.

From 2000 to 2015, the agricultural sector and food products consistently ranked among the top five of 34 sectors, highlighting their vital role in supporting the Vietnamese economy Additionally, textiles and mining emerged as significant manufacturing industries during this period Notably, the machinery sector experienced rapid growth in the last five years, suggesting its potential as a new economic driver for Vietnam in the future.

The analysis of the Leontief inverse from the 2015 table reveals that both the food products and agricultural sectors exhibit significant backward and forward linkage effects, with values exceeding 1 This indicates a strong demand for outputs from these sectors by others, while also highlighting their influence on various sectors through the procurement of intermediate goods Historically, food products have consistently emerged as the most impactful sector regarding backward linkage effects in the years 2000, 2005, and 2010, while the agricultural sector has been the most sought-after in terms of forward linkage effects during the same period.

Our analysis revealed that the significant output of food products and the agricultural sector was primarily driven by shifts in final consumption demand Additionally, the textiles sector's enhanced status within the Vietnamese economy can be linked to rising final export demand Conversely, the machinery sector experienced rapid growth over the last five years of the study period.

Over the past five years, the wholesale and retail sector has faced a significant decrease in final consumption, investment, and export demand In contrast, changes in final investment and export demand, along with technological advancements in intermediate inputs, have influenced other sectors.

Food products and the agricultural sector are vital to Vietnam's economy; however, this dependence poses challenges for economic development Relying heavily on agriculture, which often exhibits diminishing returns to scale (DRS), can hinder progress and keep the country vulnerable Many developing nations have shown that such reliance can perpetuate poverty due to limitations in product expansion without technological advancements or innovations To foster economic growth, Vietnam should prioritize the development of additional manufacturing industries, such as machinery or textiles.

1 See Chang(2003) and Reinert (2007) on the historical and structural causes of poverty

THE LINK BETWEEN FINANCIAL LEVERAGE AND

Introduction

The relationship between capital structure and investment decisions is a key focus in corporate finance, with extensive research exploring the determinants of these decisions Macroeconomic factors such as the real exchange rate, inflation, and capital flows significantly influence investment choices (Binding and Dibuasu 2017; Atella 2003; Chen Fei et al 2019) Additionally, firm-level factors, including accounting quality, financing constraints, management characteristics, and capital structure, also play a crucial role in shaping investment decisions (Myers 1977; Lang et al 1996; Gomes 2001; H.T Trinh et al 2017; Xuan Vinh Vo 2018; Shu-Miao & Chih-Liang 2017; Sang-Min Cho & Sun-A Kang 2017).

Financial leverage, defined as the ratio of total debt to total assets, indicates that higher debt levels correspond to greater financial leverage This concept is crucial for corporate investment strategies in markets characterized by transaction costs and asymmetric information This study explores the relationship between financial leverage and investment decisions among Vietnamese SMEs, utilizing a comprehensive unbalanced panel dataset from 2011 to 2015 Advanced econometric techniques, including Logit, Tobit, and Fractional logit models, are employed to analyze the data, revealing significant insights into the correlation between financial leverage, investment choices, and financing sources.

Financial leverage serves as a key indicator of capital structure, significantly impacting investment decisions in corporate finance (H.T Trinh et al 2017) Various financial theories suggest that financial leverage may be irrelevant (Modigliani & Miller 1958) or negatively correlated with a firm's investment (Myers 1977; Lang et al 1996; Aivazian et al 2005; Gome, 2001).

In their 1958 capital structure theory, Modigliani and Miller asserted that in a perfect market, a company's capital structure does not influence its investment behavior or market value, which instead relies on profitability, cash flow, and net worth However, the presence of market imperfections, such as moral hazards and information asymmetry, complicates this view Their original theory overlooked crucial factors like taxes, transaction costs, and bankruptcy costs, leading to the development of alternative theories such as the trade-off theory and pecking order theory These theories suggest that firms have preferences for different financing types, with Myers (1984) highlighting the trade-off between the tax advantages of debt and the associated bankruptcy risks.

Empirical literature challenges the leverage irrelevance theory and supports pecking order theories, indicating that firms prioritize financing sources based on cost due to asymmetric information For instance, Lang et al (1996) found that firms with higher debt ratios are less likely to seize growth opportunities compared to those with lower debt ratios, leading to reduced investment despite available growth potential Additionally, Aivazian (2005) highlighted that financial leverage negatively impacts investment decisions in Canadian publicly traded companies, particularly in firms with low growth opportunities However, these findings predominantly stem from developed countries, raising questions about their applicability to developing nations.

A study by Xuan Vinh Vo (2018) examined Vietnamese SMEs listed on the Ho Chi Minh Stock Exchange from 2006 to 2015 and found that high levels of debt significantly limit corporate investment opportunities.

Phan Q T (2018) examined the relationship between firm investment and debt financing using data from the Ho Chi Minh and Ha Noi Stock Exchanges from 2010 to 2016, finding that higher levels of debt negatively affect firm investment This suggests that an increased debt burden in a firm's capital structure leads to reduced investment levels However, it is important to note that previous studies have primarily focused on larger listed firms, leaving a gap in research regarding small and medium-sized enterprises in Vietnam.

Our research highlights a significant positive correlation between financial leverage and investment decisions among SMEs in Vietnam, aligning with H.T Trinh's 2018 study We found that higher financial leverage facilitates access to external financing for new investments, primarily because Vietnamese SMEs often face challenges with internal funding and limited external financing options, mainly reliant on bank loans Consequently, firms with elevated financial leverage are better positioned to secure credit for new investments compared to those with lower leverage This study enhances the understanding of the interplay between corporate finance and investment strategies in emerging markets, offering valuable insights for SMEs and policymakers to improve credit accessibility through strategic planning, diversifying funding sources, and minimizing information asymmetry with financial institutions.

This article is structured into several key sections: Section 2 offers an overview of small and medium-sized enterprises (SMEs) in Vietnam, Section 3 details the data and empirical methodology used, Section 4 discusses the empirical results along with an analysis, and Section 5 concludes with final remarks.

Data and Methods

This study explores the connection between financial leverage and investment decisions by analyzing quantitative surveys of small and medium-sized enterprises (SMEs) conducted in 2011, 2013, and 2015 by the Central Institute for Economic Management (CIEM) and the Institute of Labour Science.

The survey, conducted by ILSSA, DERG at the University of Copenhagen, and UNU-WIDER, encompassed a wide range of participants, gathering valuable insights into economic development.

A survey conducted across 2,500 enterprises in nine provinces in Vietnam, including Ho Chi Minh City, Long An, Khanh Hoa, Lam Dong, Nghe An, Quang Nam, Hanoi, Hai Phong, and Phu Tho, revealed consistent findings as the questionnaire remained largely unchanged (UNU-WIDER, 2018).

Between 2011 and 2013, a comprehensive survey was conducted featuring 132 questions focused on various aspects of enterprises, including their characteristics, employment, operations, costs and revenues, production, credit and loans, and environmental expenses In 2015, the questionnaire underwent minor adjustments, particularly in the areas of credit and finance It aimed to gather detailed information about firm characteristics and performance, encompassing owner attributes, workforce size, revenue and cost structures, input factors, economic constraints, and investment activities.

The surveyed enterprises span 18 sectors, including food processing, fabricated metal products, and wood manufacturing, selected from the General Statistics Office of Vietnam (GSO) The firms consist of private, collective, limited liability, joint stock enterprises, and partnerships, all officially registered under the Law of Enterprises at the provincial level A stratified sampling technique was employed to ensure representation of all enterprise types across each province.

This analysis utilizes an unbalanced sample of 6,057 micro enterprises, gathered across three survey rounds: the first in 2011 with approximately 2,512 enterprises, the second in 2013 with 2,542 enterprises, and the final round in 2015 with 2,648 enterprises Some variables were excluded and certain observations were omitted due to missing information, leading to the final total of 6,057 enterprises in the dataset.

Table 7 presents a summary of the survey data statistics and the variables utilized in our study, with a primary focus on financial leverage, defined as the ratio of debt to total assets of the enterprises from the previous survey round Additionally, we incorporated other influential variables that may affect firm investment behavior, including firm size, revenue growth, and profitability.

This study examines 33 physical assets and ownership, building on previous research (H.T Trinh et al 2017; Dang 2011) Key firm-level variables, including total assets (SIZE) and physical assets (FIXED), serve as collateral and reflect borrowing capacity from financial institutions Additionally, growth potential (GRR) and profitability (GROPF) are critical factors influencing investment decisions and financing choices The ownership status (OWN) is also considered, as it may affect management decisions, particularly in family-owned firms.

Table 7: Definition and summary statistics of the variables

Code Variables Description/Calculation method Number of

Dummy variable for new investment (=1 if the firm made a new investment during the past two years; 0 otherwise) 6,057 0.5418 0.4982

Share of external financial sources for new investment financed by bank loans and other sources that charge interest 3,282 0.412 0.4466

Share of internal financing sources or borrowing from family and friends without interest 3,282 0.5851 0.4473

Share of internal financing sources for new investment financed by borrowing from family and friends without interest 3,282 0.192 0.3462

Share of internal financing sources for new investment financed by retained earnings 3,282 0.393 0.44

Ratio of total debt to total assets at the end of the year of the previous survey round 6,057 0.0714 0.1797

SIZE Size Log of the total assets at the end of the year of the previous round 6,057 7.1718 1.7712

Growth of revenue=log of revenue in the second year minus the log of revenue in the first round of the survey 6,057 -0.0159 0.2763

Log of gross profit/revenue at the end of the year of the previous survey round 6,057 5.3614 1.4506

Ratio of physical assets (such as plants and machinery) to total assets at the end of the year of the previous survey round 6,057 1.4124 27.1834

OWN Ownership 1 if family ownership; 0 otherwise 6,057 0.6217 0.4849

ND LOCATION 1 if the enterprise is located in North

CD LOCATION 1 if the enterprise is located in

SD LOCATION 1 if the enterprise is located in South

Table 7 indicates that 54% of SMEs made new investments over a three-year period, with external financing sources accounting for 41% and internal sources, including borrowing from family and friends and retained earnings, comprising 59% A minor variation in financing sources was noted across the three periods (see Appendix 1) According to a survey conducted in 2011, internal financing was the primary source for investment.

In a survey conducted in 2013, internal financing sources, including retained earnings and borrowing from family and friends, accounted for 62% of total investments, while external financing sources made up 36% However, by 2015, the trend shifted, with internal financing surpassing external sources by approximately 20%, indicating a preference among enterprises to invest using their own capital rather than relying on bank loans.

This study aims to explore two key objectives: first, to assess how financial leverage influences the investment decisions of small and medium-sized enterprises (SMEs) in Vietnam, and second, to examine the impact of financial leverage on the sources of financing utilized for new investments.

During the first stage of the estimation, we attempt to evaluate the effect of financial leverage on the investment decisions of SMEs by estimating the following equation:

In the analysis of small and medium-sized enterprises (SMEs), the dependent variable 𝐼𝑁𝑉 𝑖,𝑡 indicates whether SME i has made a new investment during survey round t, with a value of zero if no investment occurred Financial leverage, represented by 𝐿𝐸𝑉 𝑖,𝑡−1, is calculated as the debt ratio, defined as the total debt of the SME at the end of the previous year divided by its total assets.

𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑒𝑡 𝑖,𝑡−1); 𝑋 𝑘,𝑖 refers to a vector of other control variables expected to affect the decision to invest; and 𝜀 𝑖 is the error term

In our analysis, we employ a logit model to estimate the binary dependent variable, aligning with established research by Hall et al (2000), H.T Trinh et al (2017), and Dang (2011) We incorporate several enterprise characteristics, including revenue growth, profitability, ownership structure, physical asset ratio, and geographical location, as key explanatory factors influencing SME investment decisions.

The coefficient of LEV in the model is crucial as it reflects the impact of financial leverage on a company's investment decisions Additionally, SIZE is determined by taking the logarithm of the total assets at the end of the period.

In the previous survey round, 36 firms were analyzed, focusing on key metrics such as revenue growth (GROWTH), calculated as the logarithmic difference between the second and first year's revenues Profitability (PROF) and the ratio of physical assets to total assets (FIXED) were also examined, along with ownership structure (OWN), which indicates whether the enterprise is family-owned.

ND, CD, and SD represent location dummies indicating North Vietnam, Central Vietnam and South Vietnam

Results and discussion

The results of our study regarding the decision to invest and the choice of financing sources are shown in Tables 8-13

Table 8: Investment decisions of SMEs in Vietnam (2011-2015)

Note: Values reported in parentheses are the robust standard errors (SE);

*,**, and *** indicate significance at the 10%, 5% and 1% levels, respectively

Table 9: Investment decisions of SMEs in Vietnam across three survey rounds (2011, 2013, and 2015) (cont.)

Note: Robust standard errors are shown in parentheses *** represents significance at the 1% level ** represents significance at the 5% level * represents significance at the 10% level

The findings in Tables 8 and 9 indicate a significant positive correlation between financial leverage (LEV) and investment decisions, revealing that SMEs with higher financial leverage are more inclined to pursue new investments Specifically, a one-unit increase in LEV correlates with a 7.56 increase in the likelihood of seeking new investments from 2011 to 2015 While this observation contradicts traditional finance theories and previous empirical studies, it aligns with research conducted by H T Trinh in 2017 This phenomenon may be attributed to the limited internal financing available to Vietnamese SMEs, which forces them to rely on external funding sources However, these firms often face challenges in securing external financing due to insufficient collateral and asymmetric information Consequently, financial leverage serves as an indicator of a company's creditworthiness, suggesting that SMEs with greater financial leverage are more likely to gain the trust of credit institutions and subsequently invest in new opportunities.

Revenue growth (GRR) and profitability (GRPROF) are significant factors influencing investment decisions in SMEs A positive correlation indicates that SMEs with higher revenue growth are likely to invest more, although this trend was less evident in 2013 Additionally, profitable firms tend to allocate more resources toward fixed assets While the relationship between total assets (SIZE) and investment decisions was initially unclear, strong evidence emerged when including a location dummy variable, highlighting a robust association for SMEs in South Vietnam.

The study revealed a negative relationship between the fixed asset ratio (FIXED) and investment decisions, which contradicts earlier findings by H.T Trinh et al (2017), although the clarity of this result diminished in 2011 Additionally, the collateral demanded by credit institutions for loans was found to negatively impact investment decisions Furthermore, a negative correlation exists between ownership (OWN) and the investment strategies of firms, indicating that SMEs owned by households are less motivated to pursue new investment opportunities.

Our study is also interested in the different investment decisions made by enterprises located in the following three regions of Vietnam: North Vietnam (including

In Vietnam, firms in North regions such as Ha Noi, Phu Tho, and Hai Phong demonstrate a higher likelihood of making new investments compared to those in Central regions like Quang Nam, Quang Ngai, and Nghe An, as well as South regions including Ho Chi Minh City, Khanh Hoa, and Lam Dong The analysis reveals a significant positive coefficient for new investments (ND) and a significant negative coefficient for existing investments (SD), indicating a strong investment trend in Northern Vietnam.

This study's second estimation examines the impact of financial leverage on a firm's decision to utilize either internal or external financing for new investments The findings, presented in Tables 10-13, detail the estimated results from Tobit and fractional logit models analyzing external financing sources (EXT) and internal financing sources (INT) across surveys conducted in 2011, 2013, and 2015, as well as for the full sample.

The findings indicate that financial leverage (LEV) positively and significantly influences external financing sources (EXT) while negatively impacting internal financing sources (INT) across both models LEV serves as an indicator of a firm's access to external financing, such as loans from banks or credit institutions Consequently, small and medium-sized enterprises (SMEs) with higher LEV are more likely to utilize external financial resources for new investments, whereas those with lower LEV tend to depend on internal financing This underscores the notion that elevated LEV can enhance access to external financial options.

Financial leverage significantly influences the financing decisions of SMEs, particularly in two scenarios Firstly, credit institutions often hesitate to extend additional loans to SMEs exhibiting high financial leverage due to perceived financial risks As a result, these SMEs may find it challenging to secure further credit, limiting their growth opportunities.

High financial leverage in SMEs often leads them to seek external financing sources, while those with lower leverage tend to rely on internal financing Our findings indicate that the reliance on internal financing is more prevalent, aligning with the pecking order theory proposed by Myers and Majluf (1984), which suggests that businesses prioritize their financing options based on the cost and availability of funds.

Small and medium-sized enterprises (SMEs) prioritize internal financing for investments, utilizing internal sources first Once these resources are exhausted, they turn to debt financing, and when further debt issuance is impractical, they seek equity financing.

The findings indicate a significant positive relationship between total assets (SIZE) and external financing sources, while a negative relationship exists between total assets and internal financing sources, aligning with prior research (Bhaird and Lucey 2010; Sogord-Mira 2005) Total assets serve as collateral for firms, enhancing their likelihood of obtaining bank credit Consequently, larger firms tend to favor external financing, whereas smaller firms predominantly depend on internal financing sources.

The coefficient of profitability (PROF) shows a significant positive association with external financing sources and a negative association with internal financing sources in the full sample and the 2011 survey, challenging the pecking order theory and earlier research; however, these associations were not significant in the 2013 and 2015 surveys While prior studies indicate a negative relationship between external financing use and firm profitability, our findings suggest that profitable firms still opt for external financing Additionally, there is insufficient evidence linking revenue growth (GRR) or family ownership (OWN) to external financing choices, as their coefficients are statistically nonsignificant Notably, a significant positive correlation exists between revenue growth and retained earnings, along with a significant negative correlation with other financing sources in the 2011 and 2015 surveys and the overall sample.

Table 10: The relationship between financial leverage and the choice of financing sources (2011-2015)

Tobit Flogit Tobit Flogit Tobit Flogit Tobit Flogit

Note: Standard errors are shown in parentheses

*** represents significance at the 1% significance level, ** represents significance at the 5% significance level, * represents significance at the 10% significance level

Table 11: The relationship between financial leverage and the choice of financing sources (2011)

Conclusion

This study examines the impact of financial leverage on investment decisions and financing sources in Vietnamese small and medium-sized enterprises (SMEs) using survey data from 2011 to 2015 Contrary to established theories by Myers (1977) and Gomes (2001), as well as empirical findings by Lang et al (1996) and Aivazian et al (2005), which suggest that high debt ratios limit growth opportunities and investment likelihood, our results indicate a positive relationship between financial leverage and investment decisions in Vietnamese SMEs This suggests that firms with higher financial leverage are more inclined to pursue new investments for expansion, challenging Modigliani & Miller's theory and highlighting that companies may have distinct preferences in their investment decision-making processes.

Firms with higher financial leverage tend to rely more on external financing sources while decreasing their use of internal financing for new investments This trend is attributed to high-leverage companies having better access to external funds due to their strong credit records In the context of Vietnamese SMEs, there is a noticeable shift from internal to external financing sources (H.T.Trinh et al 2017) Conversely, low financial leverage restricts access to external financing, leading companies to utilize internal funds for expansion As operations grow successfully, businesses enhance their ability to secure external financing, further increasing their financial leverage.

Nevertheless, in contrast to previous studies (H.T.Trinh et al 2017), our results show that firm profitability has a significant impact on the choice of external financing

48 sources because firms still rely on external financing sources even if they make high profits, suggesting that profitable enterprises invest more in fixed assets

Reducing credit constraints and enhancing financial leverage for SMEs can effectively stimulate new investments Policymakers should focus on improving credit access, alleviating information asymmetry, and providing direct subsidies to support SMEs Additionally, the government can enhance awareness of financial institutions offering investment opportunities and create information-sharing channels between investors and SMEs Lowering trade facilitation costs can attract more foreign investment, while promoting SME participation in global networks is also essential for fostering growth.

Our study acknowledges certain limitations while offering valuable insights for future research It specifically focused on SMEs in Vietnam, a representative country within emerging economies Consequently, conducting comparative studies in other developing countries would be advantageous for future investigations.

INFLUENCE OF FARMER UNION MEMBERSHIP ON THE

Introduction

Rural development is crucial for poverty reduction in low-income countries, as a significant portion of the poor population resides in these areas, relying on farming activities like rice, livestock, and agro-forestry for their household income However, these individuals often face challenges such as low education levels, limited skills, and restricted access to essential services like financial and healthcare resources Research indicates that organizing poor farmers into groups or cooperatives can help them overcome these barriers and improve their livelihoods.

Agricultural cooperatives play a crucial socioeconomic role by assisting family farms in overcoming challenges related to external economies of scale and enhancing market competitiveness, as highlighted by Valentinov (2007).

Agricultural cooperatives have been shown to significantly alleviate poverty among farmers and communities (2017), serving as vital tools for addressing rural poverty (Deriada, 1995) These organizations enhance individual negotiating power and reduce transaction costs (Bernard and Spielman, 2009; Francesconi and Ruben, 2012; Markelova et al., 2009; Valentinov, 2007) However, some studies highlight the poor performance of agricultural cooperatives in developing nations (Fischer and Qaim, 2012) This paper explores the effects of farmers’ union membership on rural households in Vietnam, comparing findings with other developing countries.

Since the doi moi renovation in 1986, Vietnam has seen a surge in the activities of various organizations, including community-based and nongovernmental organizations The farmers’ union serves as a key case study for assessing the influence of social networks on farm households due to its strong grassroots connections and substantial membership, particularly within Vietnam's unique single-party system Following doi moi, mass organizations like the farmers’ union have gained independence and play a crucial socioeconomic role, offering significant support to impoverished individuals through enhanced education, healthcare, living conditions, and access to financial services.

The Vietnamese Bank for Social Policy (VBSP) plays a crucial role in providing credit to low-income individuals who struggle to access traditional loans requiring collateral, inspired by the successful model of the Grameen Bank in Bangladesh Utilizing a group lending scheme, the VBSP relies on village heads and commune leaders to form borrower groups and monitor repayments, thereby enhancing accountability This group-based borrowing approach not only reduces transaction costs but also addresses issues related to asymmetric information Research by Hong Sun et al (2018) highlights the significant impact of social capital, particularly through kinship and friendship, on the borrowing behavior of rural households in China, with the farmers' union identified as a key player in this dynamic However, the existing empirical literature connecting social networks to farm households remains limited.

This study examines the influence of farmers' union membership on economic performance and financial matters, particularly in developing countries like Vietnam Previous research, such as Giannakis et al (2018), demonstrated that union membership positively affects off-farm work decisions Similarly, Newman et al (2014) found that being part of high-quality networks enhances saving levels Furthermore, Takashi (2009) highlighted the significant role of farmers' unions in microcredit programs in Vietnam Our research contributes to existing literature by providing empirical evidence on the importance of social networks, specifically farmers' unions, in rural household production and credit volume.

This paper is organized into several sections: Section 2 provides a literature review and an overview of farmers’ unions in Vietnam; Section 3 outlines the data and variables utilized in the study; Section 4 details the analytical methodology; Section 5 presents the empirical findings and their interpretations; and finally, Section 6 offers the conclusion.

Context and Literature Review

The Vietnam Farmers’ Union (VNFU), founded on October 14, 1930, is a sociopolitical organization led by the Communist Party of Vietnam that plays a crucial role in farmer movements and rural development Open to individuals over 18 from diverse backgrounds, the VNFU allows members from various agricultural sectors to join by voluntarily accepting its philosophy and regulations The primary motivation for farmers to join the VNFU is to gain support in developing their agricultural enterprises Recently, the VNFU has also helped farmers access credit from the Vietnam Bank for Social Policies (VBSP), aiming to provide collateral-free loans and promote savings and credit among its members.

The VNFU is involved in various national programs aimed at job creation, agricultural extension, and vocational training, providing support through 52 groups focused on managing repayment Membership in the VNFU offers economic advantages that may surpass the organization's initial goals.

The microcredit program, initiated by Professor Muhammad Yunus at Grameen Bank in Bangladesh in 1976, has served as a model for similar initiatives in developing countries, providing loans to impoverished individuals without collateral, particularly benefiting women Over seven million Bangladeshis have accessed credit through this program, which has been recognized as a vital tool for poverty alleviation and sustainable economic improvement While some studies indicate positive outcomes, such as increased household income and expenditures, particularly for women borrowers in Malaysia and minority students in Vietnam, others suggest mixed or negligible impacts on household income Research has shown that microfinance can enhance health service usage and nonfarm income, yet some findings indicate that loans are often utilized for investments rather than transformative income growth, with factors like agricultural shocks limiting potential benefits.

On the other hand, several papers have pointed out that rural credit is influenced by social institutions or groups For example, Takashi (2009) mentioned that mass

In Vietnam, 53 organizations play a crucial role in organizing consultative meetings to select and recommend candidates for credit programs Among these, the Farmers’ Association is highlighted as the most influential in facilitating bank loans Following the establishment of the VBSP loan, membership in the Farmers’ Association saw a significant increase However, empirical analysis to support this claim was not conducted Research by Newman, Tarp, and Broeck (2014) indicates that membership in organizations such as the Farmers’ Association and Women’s Union significantly affects household formal savings in rural areas of Vietnam.

Materials and Methods

4.3.1 Study sites and survey approach

The Vo Nhai district, recognized as one of the most rural regions in Vietnam, was chosen for this study due to the implementation of a rural credit program and its status as the area with the highest poverty rate in the northern mountainous region of the country.

Thai Nguyen Province, located in northern Vietnam, spans 3,533.2 km² and had a population of 1,190,600 as of 2015 Renowned for its tea industry, the province dedicates 16,000 hectares to tea production Despite its agricultural significance, the poverty rate in Thai Nguyen was 11.1% in 2014, with Vo Nhai District facing over 40% In 2016, approximately 150,860 households engaged in agriculture, forestry, livestock, and aquaculture, making up 66.7% of the province's households, compared to the national average of 53.7% Vo Nhai District notably has the highest concentration of farm households in the province at 83.15%.

In this study, a simple random sampling technique was applied From the lists of HHs provided by the district offices, total of 401 households were randomly selected

The study's final sample size consisted of 342 households, after excluding 59 due to missing information Within this sample, 111 households were members of a farmer union, while 81 were non-members, all from Vo Nhai district in Thai Nguyen Province Bao Hieu district in Hoa Binh Province was excluded from the main estimation to focus solely on the data from Thai Nguyen Province.

192 farm HHs, which is 0.5% of the total number of farm households in the selected area The primary data were collected through an HH survey conducted in 2015-2016

In our study, households (HHs) without farmers' union membership were classified as the control group, while those with membership formed the treatment group The HH survey addressed various aspects, including demographic characteristics, access to credit, and economic performance Respondents provided information on the age, education level, gender, occupation, ethnicity, and poverty status of the HH head Additionally, the survey explored HH access to resources, credit, and primary production activities For credit analysis, participants shared their loan transaction history from any sources over the past decade.

This study utilized the propensity score matching (PSM) method to assess the effects of farmers’ union membership on various outcome variables, including household economic performance indicators such as average livestock and crop production, total income and expenditures, as well as the average credit volume obtained from the VBSP program.

This study selects household (HH) variables based on a thorough literature review (Caliendo and Kopeinig, 2008; Becker and Ichino, 2002) and identifies several potential covariates from HH characteristics that may influence both treatment and outcome variables A summary of the control variables and the variables of interest, along with their underlying assumptions, is provided in Table 14.

Table 14: Description of selected variables

Family size Total members in the household Members

Sex The gender of the HH head 1= male; 0 = female

Age of household head Age of household head Years

Education of household head The number of years in school of the household head Years Occupation of household head

Dummy variable for occupation of the household head; 1 for working full-time on farm, 0 otherwise

Credit dummy Dummy variable for credit status of HH; 1 = has borrowed money from the VBSP, 0 otherwise Total agricultural and forestry land holdings

Total lands owned by the HH 1000 m 2

Livestock production contributes significantly to the overall agricultural economy, with total production valued in millions of VND Similarly, crop production also plays a crucial role, reflecting a substantial financial output in millions of VND The total monthly income of households is measured in millions of VND, highlighting the economic activity within the community Conversely, household expenditures in a month are also quantified in millions of VND, indicating the financial obligations faced by families Additionally, the credit volume from the Vietnam Bank for Social Policies (VBSP) represents the total borrowing amount in millions of VND, essential for supporting agricultural and household needs.

Analyzing the causal effects of farmers' union membership on potential outcome indicators is crucial to address endogeneity bias It is essential to account for both observable and unobservable characteristics when randomly assigning individuals to treatment groups (Wossen et al., 2017) This study utilizes Propensity Score Matching (PSM) to effectively control for these variables.

To address endogeneity bias, we utilize propensity score matching (PSM) to align treated households with similar untreated ones, as proposed by Rosenbaum and Rubin (1983) This method involves constructing counterfactual outcomes for the treatment group based on their propensity scores derived from a set of covariates By calculating the differences in outcomes between matched observations, we can estimate the average treatment effects (ATEs) and average treatment effects on the treated (ATETs) for the outcome variables, following the methodology established by Tran and Goto (2019).

In the first step, the propensity score was estimated by applying a probit model, which is expressed as follows:

Additionally, 𝐷 𝑖 =1 if an HH is a member of the farmers’ union, and 𝐷 𝑖 =0 if an HH is not a member of the farmers’ union

𝑋 𝑖 is a vector of HH characteristics

𝛷 represents the standard normal cumulative distribution function

To achieve optimal matching quality and ensure a balanced sample concerning the chosen covariates, we employed various models for estimating the propensity score, each utilizing a distinct set of variables The specific variables selected for each model are detailed in Table 3.

The propensity score was estimated using a selected probit model, followed by one-to-one nearest neighbor matching with a 0.01 caliper Treated units lacking control units within the specified caliper were excluded from the estimation, while control observations could be utilized for matching multiple times.

In the next step, we treat selected covariates as pseudo-outcome variables and estimate pseudo-ATETs based on them to check the balance of each model following

Imbens and Rubin (2015) state that effective matching should yield pseudo-ATETs that are close to zero and statistically insignificant, indicating the plausibility of the unconfoundedness assumption The model demonstrating the best balancing properties will be selected for estimating the Average Treatment Effects (ATEs) and Average Treatment Effects on the Treated (ATETs) in the subsequent analysis.

In the second step, each selected model will be used to estimate the ATEs (equation

2) and ATETs (equation 3) as follows:

ATET = E(𝑌 1 − 𝑌 0 |𝐷 = 1) = 𝐸(𝑌 1 |𝐷 = 1) − 𝐸(𝑌 0 |𝐷 = 1) (3) where 𝑌 1 and 𝑌 0 are potential outcomes, and 𝐷 denotes a treatment indicator

In our dataset, we can only observe 𝐸(𝑌 1 |𝐷 = 1), while 𝐸(𝑌 0 |𝐷 = 1) remains unobserved This unobserved term is a counterfactual, which we estimate by averaging the outcomes of the matched control group.

We utilized Propensity Score Matching (PSM) due to its effectiveness in adjusting for confounding variables in observational studies PSM offers distinct advantages over traditional regression methods, as it forms treatment and control groups with similar propensity scores, unlike regression, which estimates outcomes for unmatched households Additionally, PSM reduces the impact of self-selection bias linked to observable covariates (Rosenbaum and Rubin, 1983).

To ensure the reliability of our findings, we utilized additional matching techniques, including the inverse probability weighted (IPW) method and regression adjusted (RA) methods While propensity score matching (PSM) aligned treatment and comparison group observations based on their propensity scores, IPW estimators calculated probability weights to derive weighted averages of outcomes for each treatment level.

Results and discussion

Table 2 presents the mean differences between the selected variables of FU member households and non-member households Notably, heads of households with FU membership tend to have a higher level of education compared to those without membership.

Younger and predominantly female heads of households (HH) are more common among members compared to nonmembers Additionally, households in the control group possess a greater amount of farmland than those in the treatment group, indicating a disparity in selected covariates that points to potential self-selection issues.

The treatment households exhibit higher average values in livestock production, expenditure, and credit volume compared to the control households, with an average credit volume that exceeds the control group by 4.3 million VND However, there are no significant differences in mean values for livestock and crop production, and the treatment group shows a lower average total income than the control group Additionally, there are no statistically significant differences in other covariates between the treatment and control groups.

Participation in farmers' unions often varies among households due to differing self-selection processes The average differences in outcome variables between treatment and control groups can result in biased conclusions about the treatment effects Although Propensity Score Matching (PSM) cannot fully address self-selection bias, we identified several sets of criteria to enhance the analysis.

60 covariates to estimate the Propensity score then test the consistency of the impacts results by using different models and different matching algorithms

Table 15: The mean difference of households without farmers’ union membership (Treatment) and with farmers’ union membership

Variables Treatment Group Control Group Difference (T-C)

Obs Mean S.D Obs Mean S.D Mean S.E

Age of the HH head 111 45.585 0.964 81 48.283 1.251 -2.698 * 1.554

Educational level of the HH head

Total agricultural and forestry land holdings

Data source: The author’s calculations from the Household Survey from 2016-2017 ***, ** and * represent significance at the 1%,

5% and 10% levels, respectively; 100 USD = 2,233,000 VND (in 2016)

To assess the impacts of FU's membership, we utilized the methodology outlined in section 3.3 According to Caliendo and Kopeinig (2008), propensity score modeling is not designed to predict individual selection into treatment and control groups, but rather to achieve balance among the selected covariates Our analysis identified two models, out of eight, that yielded the most balanced samples post-matching, as detailed in Table 3 Prior to matching, significant mean differences were observed in covariates such as the age and educational level of the household head and occupation dummy, but these differences became nonsignificant after matching Additionally, for nearly all covariates, the mean differences approached zero, with the exception of total agricultural and forestry land holdings, which remained nonsignificant This indicates an improvement in sample balance.

Table 16: Balance checking before and after matching

Mean SE Mean SE Mean SE

Age of the HH head -2.698* 1.554 0.337 1.55 1.411 0.343 Educational level of the HH head 1.065*** 0.343 0.000 0.302 -0.011 0.327

Total agricultural and forestry land holdings -1.774 5.753 7.638 8.691 3.362 6.687

Note: In model 1, total agricultural and forestry land holdings was excluded in the propensity score model (probit model) In model 2, sex was excluded in the propensity score model

A caliper of 0.01 is applied for one-to-one nearest neighbor matching based on the estimated propensity score

***, ** and * represent significance at the 1%, 5% and 10% levels, respectively

The analysis of Average Treatment Effects (ATEs) and Average Treatment Effects on the Treated (ATETs) reveals that farmers' union membership significantly enhances livestock production, with member households (HHs) producing over 9.121 million VND, and up to 11.556 million VND in model 2 Additionally, member HHs benefit from higher credit volumes compared to non-member HHs, as union membership facilitates access to substantial credit for production improvements This trend is consistent across various models and matching methods, underscoring the positive impact of farmers' union membership on both livestock production and credit access Our findings align with previous research highlighting the critical role of social factors, such as farmers' union membership, in supporting rural households and microcredit initiatives, particularly following the introduction of VBSP loans in 1996 Thus, the study reinforces the notion that farmers' unions are instrumental in boosting loan availability and livestock output.

Additionally, the ATET estimator for the outcome variables reveal that farmer’s union membership has a positive effect on livestock production and credit volume

The findings indicate that farmers' union membership plays a significant role in household production, echoing the ATE results However, there is a discrepancy between the two models analyzed: model 2 shows a significant effect of membership on livestock production, while model 1 does not Additionally, the ATET results reveal a significant influence of membership on credit volume in model 1, contrasting with a nonsignificant impact in model 2 Furthermore, the ATET results exhibit inconsistencies across various matching methods.

We found no evidence of an impact of FU membership on other potential outcomes such as crop production, total income and HH expenditures

Our analysis reveals that both Propensity Score Matching (PSM) and linear regression confirm the statistical significance of farmers' union membership on household livestock production and credit volume, even after accounting for various covariates.

The study reveals that membership in farmers' unions significantly influences household credit volume and livestock production, while its effect on crop production is inconsistent This indicates that being part of a farmers' union offers substantial economic advantages for its members and their households It's important to note that union members are not leveraging their position to secure preferential credit; rather, their higher education levels contribute to their perceived creditworthiness Furthermore, the correlation between education level and union membership status plays a crucial role in shaping credit utilization for household production.

Table 17: The results of PSM estimation

ATE SE ATE SE ATET SE ATET SE

A caliper of 0.01 is applied for one-to-one nearest neighbor matching based on the estimated propensity score

***, ** and * represent significance at the 1%, 5% and 10% levels, respectiv

Table 18: Results of ATE estimation: Impact of membership on HHs

Membership (1 vs 0) PSM NN-MATCH IPW RA

Note: In this PSM estimation, all the covariates (family size, sex, age, education level, occupation dummy, credit dummy, total land) are included when estimating the propensity score

Table 19: Results of ATET estimation: Impact of membership on HHs

Membership (1 vs 0) PSM NN-MATCH IPW RA

Note: In this PSM estimation, all the covariates (family size, sex, age, education level, occupation dummy, credit dummy, total land) are included when estimating the propensity score

Conclusion

This study investigates the role of farmers' union membership in enhancing production and access to microcredit for farm households in rural Vietnam, contrasting previous literature focused primarily on microcredit's impact on household welfare Conducted in Vo Nhai district, Thai Nguyen province—an area with a notably high poverty rate—the research employed propensity score matching (PSM) and linear regression to ensure result consistency Findings indicate that membership in farmers' unions significantly boosts livestock production and increases credit availability for rural households.

In particular, the ATE estimator results show that membership in farmers’ unions is positively impacts to the livestock production and credit volume of member HHs

Households (HHs) that are members of farmers' unions tend to secure higher credit volumes and achieve greater livestock production compared to non-member HHs, demonstrating a clear and significant impact on credit access Although the results from the Average Treatment Effect on the Treated (ATET) estimator were inconsistent, they suggest that better-off HHs, in terms of credit and production, are more inclined to join farmers' unions This indicates that farmers' unions offer potential benefits to farming households, a notion that, while seemingly obvious, has not been thoroughly explored in previous studies The findings highlight that membership in farmers' unions correlates with a more substantial increase in livestock production, supporting the idea that rural households pursue specific objectives through collective action.

The difference in the results originates from a difference in the matching method and the different PSM models This discrepancy might be a limitation of this study’s methodology

In conclusion, our estimates indicate that farmers’ union membership has significant impacts on HHs’ production and credit volume Therefore, the economic

Membership in farmers' unions offers 69 benefits that can significantly enhance household production beyond the organization's primary goals This research provides empirical evidence highlighting the crucial role of farmers' unions in Vietnam's rural development, reinforcing the importance of social organizations The findings can assist policymakers in understanding the positive impacts of local organizations on agricultural progress.

The farmers' union should strengthen its initiatives to improve access to resources, particularly financial services Additionally, exploring the interactions between local organizations and rural households across various regions can inform potential policies and support for social organizations involved in rural development in Vietnam.

SUMMARY OF THE STUDY

Tiêu đề	Essays on financial activities and industrial change: The study case in Vietnam
Tác giả	Nguyen Thi Bich Ngoc
Người hướng dẫn	Prof. Ichihashi Masaru
Trường học	Hiroshima University
Chuyên ngành	International Development and Cooperation
Thể loại	Doctoral dissertation
Năm xuất bản	2020
Thành phố	Hiroshima

Định dạng
Số trang	116
Dung lượng	2,15 MB