Introduction
Corporate Income Tax (CIT) plays a crucial role in influencing the growth of businesses and the overall economy, particularly in developing countries like Vietnam, where fiscal pressures pose significant challenges The aging population increases the strain on pension and healthcare systems, while the declining proportion of youth under 14 further complicates economic development Additionally, trade liberalization stemming from free trade agreements with the EU and WTO is expected to reduce government revenue from tariffs in the future To alleviate these fiscal pressures, it is essential to enhance corporate tax revenues by fostering the establishment of new firms and improving the productivity of existing ones.
Vietnam has implemented a declining corporate income tax (CIT) system to enhance its competitive investment environment, similar to other countries in the region Prior to 2000, foreign firms were subject to a 25% tax rate under the Foreign Investment Law, while domestic firms faced a higher rate of 32% under the Domestic Investment Law In 2004, Vietnam unified these laws into a general Investment Law, allowing both foreign and domestic investments to operate under the same framework However, the CIT rates remained different, with the new law setting the CIT at 28% for domestic firms while maintaining the 25% rate for foreign firms, effective from January 1, 2004.
In December 2004, the Prime Minister issued a decision to carry out
The "Tax System Reform Program 2005-2010" aimed to enhance production, boost exports, attract investment, and foster technological innovation, ultimately driving sustainable economic growth and improving living standards Implemented during a time of global technological advancement and market openness, the program sought to align Vietnam's tax policies with international trends of reducing tax rates, particularly for Corporate Income Tax (CIT) By 2008, the program aimed to unify tax rates and incentives across various economic sectors to promote investment and ensure fair competition Following Vietnam's accession to the WTO in 2007, the reform established a competitive environment for both domestic and foreign businesses, culminating in a standardized tax rate of 25% in January 2009, thereby eliminating discrimination between different business entities.
The corporate tax cut for domestic firms in Vietnam has enhanced their competitiveness by aligning tax rates with those of foreign firms, creating a favorable tax environment compared to several Asian countries Notably, Singapore reduced its corporate income tax (CIT) to 22% in 2003, followed by a further decrease to 20% in 2005 Similarly, Malaysia adjusted its tax rate to 25% in 2009, while China initiated a significant tax reform, lowering its rate from 33%—which had been in place since 1997—to 25%.
On December 6, 2004, Decision No 201-2004-QD-TTg was issued to reform the tax policy system for the period of 2005-2010 This decision introduced several new taxes, including anti-dumping, anti-subsidy, anti-discrimination, environmental protection, property, and land use taxes Additionally, it called for revisions and supplements to existing taxes, particularly the corporate income tax.
2007 Thailand and Indonesia still kept theirs at 30% and 28% respectively in 2009.
In May 2011, Vietnam implemented the "Tax System Reform Strategies 2011-2020," aiming to reduce corporate income tax (CIT) to attract investment and enhance firms' financial capabilities This strategy focuses on promoting investments in high value-added sectors, including high technology and biotechnology, while also supporting areas with challenging socio-economic conditions As a result, the CIT law was revised in 2013, lowering the tax rate from 25% to 22% in January, thereby improving competitiveness and encouraging further investment.
In 2014, Vietnam reduced its corporate tax rate to 20% for firms with revenue above 20 billion VND, aligning with regional trends to attract investment Since July 2013, firms with revenue equal to or less than 20 billion VND also benefited from this 20% tax rate This move was significant as neighboring countries like Thailand lowered their tax rates from 30% to 20% between 2000 and 2013, while Singapore reduced its rate to 17% in 2010 and Malaysia set it at 24% in 2015 In contrast, countries such as China and Indonesia maintained their tax rates at 25% during this period.
The figure (1.1) shows the corporate tax rates in Vietnam during the period 2006-2016 From 2009 onwards, domestic and foreign firms have the same tax rate.
This paper examines the 2009 reform of the Corporate Income Tax (CIT), which established equal tax treatment for domestic and foreign firms for the first time We analyze the impacts of this policy change on domestic firms in comparison to foreign firms, focusing on capital stock, investment, employment, and labor income Additionally, we explore how these changes are associated with various firm characteristics, aiming to answer key questions regarding the implications of the CIT reform.
First, whether or not domestic firms accelerated investment and capital stock expansion due to the corporate tax cut policy in 2009 Although a couple
2 Data is taken from OECD website: https://www.oecd.org/tax/beps/corporate-tax-statistics- database.htm
Figure 1.1: Corporate tax rates in Vietnam
Domestic firms foreign firms All firms
Research on corporate tax rates in Vietnam has primarily focused on foreign direct investment and qualitative studies, with a notable lack of quantitative analysis regarding the effects of the 2009 tax cut policy on firm behavior Understanding how this policy influences capital stock and investment is crucial, as empirical evidence suggests that favorable tax policies can stimulate investment and enhance capital stock, as demonstrated by studies such as Federici and Parisi (2015) and Djankov et al (2010) Therefore, it is anticipated that similar positive outcomes will occur in Vietnam.
To tackle this issue, we utilize a Difference-In-Difference methodology, which simulates an experimental research design by comparing the effects of a treatment on a treated group versus a control group Specifically, we examine the 2009 tax cut policy, which primarily benefits domestic firms, although some are ineligible due to variable common CIT rates, while foreign firms do not qualify for the tax cut.
The 2009 tax cut policy significantly impacted the economy by fostering job creation and enhancing employee incomes Research indicates that this tax reform led to positive changes in firm behavior, ultimately benefiting both the economy and workers By analyzing the effects of the tax cut on employment rates and wage growth, it becomes evident that the initiative played a crucial role in stimulating economic activity and improving financial conditions for employees.
Previous studies by Nguyen and Hua (2013), Oxfam (2016), and OECD (2018) indicate that salary assessments in Vietnam remain incomplete To explore the effects of this issue, we will utilize a similar methodology, focusing on employment rates and labor income growth as our new dependent variables.
The responsiveness of firms to tax policy may vary based on their characteristics, such as size and liquidity ratio Understanding these differences is crucial, as they can influence a firm's attitude toward tax regulations Additionally, analyzing the treatment effects across various economic sectors will provide insights into which sectors are benefiting from changes in tax policy.
This paper is structured as follows: Section 2 reviews the relevant literature, while Section 3 outlines the econometric methods employed In Section 4, the dataset and variable measurements are detailed Section 5 presents the regression analysis results and robustness checks for four key dependent variables: capital stock, investment rate, employment, and labor income growth Section 6 explores potential extensions, suggesting that the treatment effect may vary with certain variables Finally, Section 7 provides a conclusion.
Literature Review
The effect of corporate tax policy on capital stock and investment are some of the main debates of the academia and researchers Hall and Jorgenson
In their foundational works, McFadden (1967) and Hall and Jorgenson (1969) explored the relationship between capital accumulation and investment through the lens of neoclassical theory They concluded that, under static expectations, current investment is influenced by the existing capital stock and the historical gap in desired capital levels Additionally, they found that direct taxes elevate the user cost of capital, which reduces the desired capital stock and subsequently hampers investment However, the reliance on static expectations presents a limitation in their analysis.
Chirinko (1986) enhanced Jorgenson’s investment theory by incorporating expectations, positing that investment decisions are influenced by both current circumstances and anticipated future conditions Firms are motivated to invest when the expected marginal revenue product of an additional capital unit aligns with the anticipated marginal cost of capital The optimal capital stock is positively correlated with expected real output and negatively correlated with the anticipated user cost of capital Expectations, which are not directly observable, are derived from historical data for each variable, leading to net investment being influenced by both current and past changes in desired capital stock Despite Chirinko's new investment function formulation, the impact of taxes on investment and capital stock remains consistent with Jorgenson’s original findings.
Sandmo (1974) analyzed the effects of corporate income tax on capital accumulation and composition by developing a model that incorporates both short-term and long-term capital goods His findings indicate that corporate income tax distorts pre-tax factor pricing, leading to higher rental rates relative to wages, which encourages firms to opt for less capital-intensive production methods and substitute capital with labor Additionally, Sandmo emphasized the significance of depreciation allowances in investment decisions, noting that a constant depreciation allowance under corporate income tax results in a capital stock composition that favors more durable capital goods.
The empirical analysis reveals a divergence in findings regarding the impact of corporate taxes on investment and capital stock, leading to two distinct schools of thought Some studies indicate a significant negative effect of corporate taxes on investment, while others find no correlation between the two factors.
Federici and Parisi (2015) analyzed the impact of corporate taxation on investment in Italy using a firm-level panel dataset spanning 12 years (1994-2006) They noted significant changes in the total corporate tax rate, which decreased from 53.95% in 1994 to 41.25% in 1998, and further to 33% in 2004 Utilizing an investment equation based on the Euler equation from the standard neoclassical accumulation model, the authors applied the Generalized Method of Moments (GMM) for their estimates Their findings revealed that a one-percentage-point reduction in the average tax rate correlates with a 12% increase in the investment rate.
Dwenger (2009) examines the short-term and long-term impacts of corporate income tax on investment and capital stock in Germany from 1987 to 2007 Utilizing an Error Correction Model, the study finds that a 10% increase in the user cost of capital results in a 13% decrease in capital stock over the long term, with an immediate short-term reduction of 5% Following the 2008 Corporate Tax Reform in Germany, which lowered the tax rate from 25% to 15%, the user cost of capital decreased With an elasticity estimate of -1.29, the author predicts that this tax reform will lead to a long-term capital stock increase of approximately 0.11 percent However, the modest rise in capital stock from the significant corporate tax reduction is attributed to stringent depreciation allowances that diminish the overall tax impact.
In a country level analysis, Djankov et al (2010) measured the effects of corporate income taxes on investment and entrepreneurship in 85 countries in
In 2004, a significant adverse effect was observed, with a 10% increase in tax leading to a 2% decrease in the aggregate investment to GDP ratio, a 1.4% decline in the entry rate, and a reduction of 1.9 firms per 100 people in business density Notably, while tax rates negatively affected investment in the manufacturing sector, the service sector, which is characterized by a high prevalence of informal activities, remained largely unaffected by these tax changes.
Cerda and Larrain (2010) conducted a study on the impact of corporate tax on labor and capital demand in Chile from 1896 to 1996, analyzing 970 manufacturing firms Their research revealed a negative correlation between tax rates and both labor and capital, with the effect on labor being nearly twice as significant When firm size was considered, smaller firms demonstrated a greater sensitivity to tax impacts on capital, while larger firms showed more sensitivity regarding employment This suggests that tax rates directly influence the availability of internal funds, which are more critical for smaller, credit-constrained firms In contrast, Egger et al (2018) found that corporate taxation affects managerial firms, typically larger ones, more than entrepreneurial firms, which are often smaller and financially constrained Their analysis, grounded in the Agency model, indicates that high corporate taxes lead to reduced investment as firms prioritize dividend payouts to leverage tax-deductible debt, consequently diminishing internal funds As a result, entrepreneurial firms, which rely more on external financing, are less sensitive to investment changes compared to managerial firms with substantial internal resources.
Bustos et al (2004) argue that higher corporate tax rates do not necessarily lead to a reduction in desired capital stock They suggest that depreciation allowances and interest deductions can offset the impact of increased taxes by lowering taxable profits When these financial incentives are substantial, high taxes may actually reduce the cost of capital, potentially maintaining or even increasing capital stock Utilizing neoclassical theory, the authors developed an estimation equation incorporating the capital-output ratio, investment-capital ratio, user cost of capital, and long-run elasticity of substitution between capital and labor for Chilean firms from 1985 to 1995 Their findings indicate that the overall effect of taxation on capital stock is minimal.
Research by Ljungqvist and Smolyansky (2018) highlights the negative effects of corporate taxes on employment and income growth, revealing that a one percentage point increase in corporate tax rates can lead to a 0.2 percent decrease in employment and a 0.3 percent decline in total wage income at the county level in the US from 1969 to 2010 Similarly, Shuai and Chmura (2013) analyze panel data from 1990 to 2012 at the state level, confirming a significantly adverse impact of corporate taxes on employment growth Although corporate tax cuts can initially stimulate job creation, this effect tends to be temporary, primarily manifesting in the first year following the tax reduction.
Yatskovskaya (2012) investigates the impact of tax credits on employment in the US using industry-level data from 1999 to 2008 Employing panel data regression with fixed-effects and between-effects models, the study incorporates control variables such as total assets, total deductions, and net worth The findings reveal no significant evidence that tax credits influence employment levels These results contrast sharply with the conclusions drawn by Ljungqvist and Smolyansky (2018).
Feld and Kirchgößner (2003) examined the effects of corporate and personal income taxes on firm location and employment in Switzerland, utilizing a panel dataset from 26 Swiss cantons between 1981 and 1991 for firm distribution analysis and from 1985 to 1997 for employment analysis They employed a Seemingly Unrelated Regression estimator to address the simultaneity between dependent and explanatory variables Their findings indicate that both corporate and personal income taxes influence firms' decisions to relocate, ultimately leading to a reduction in employment within the cantons.
Research on the relationship between corporate tax, investment, and capital stock in Vietnam is limited, primarily consisting of qualitative studies and basic descriptive statistics Additionally, there is a scarcity of analyses examining the effects of corporate tax on employment and labor income within the country A notable survey conducted in 2011 by the United Nations Industrial Development Organization (UNIDO) and the Vietnam Ministry of Planning and Investment surveyed 1,426 manufacturing firms across several major provinces in Vietnam.
A recent survey conducted in key regions of Vietnam, including Ho Chi Minh City, Binh Duong, Dong Nai, Hanoi, Vinh Phuc, Bac Ninh, Da Nang, and Ba Ria-Vung Tau, examined factors influencing investment decisions among both domestic and foreign firms According to a UNIDO report from 2014, taxation ranks as the third most significant factor affecting these decisions, following economic and political stability, with labor costs and the legal framework of the country also playing crucial roles Additionally, Oxfam International has contributed to the analysis of Vietnam's tax policies through various reports on tax incentives, offering recommendations for improvement However, these reports primarily present survey data and statistical descriptions without establishing a definitive model to demonstrate the relationship between corporate taxes and investment in Vietnam.
Recent studies on corporate tax in Vietnam have primarily focused on foreign direct investment (FDI), rather than domestic investment Nguyen and Hua (2013) assessed the tax burden on FDI from 1999 to 2011, incorporating labor costs and inflation as control variables Their findings indicated a negative relationship between tax rates and total implemented FDI capital However, the study lacked a firm-level analysis, which limits its ability to account for the varying characteristics of individual firms within the model.
Econometric Approach
We employ a difference-in-difference (DID) methodology to evaluate the impact of the 3% corporate income tax reduction implemented in 2009 on corporate behavior This approach allows us to assess the effects of a policy change by comparing treated groups, which experienced the tax reduction, to control groups that did not By analyzing the differences in outcomes before and after the tax change, we can effectively measure the treatment effect on firms.
In this analysis, we consider a policy change occurring at time k, with t0 representing the time before the change and t1 indicating the time after We define the treatment status of individual i at time t as U, where individuals in the treated group are denoted by di=1, and those in the control group by di=0 An individual is classified as part of the treated group if their treatment status at time t1 is equal to 1; otherwise, they belong to the control group.
According to Blundell and Costa Dias (2009), an individual's outcome (yit) can be expressed as yit = α + δidit + uit, where E[uit|di, t] = E[ni|di] + mt In this equation, ni represents the unobservable individual fixed effect, while mt signifies an aggregate macroeconomic shock.
α+E[δi|di=1] +E[ni|di=1] +mt i f di=1&t=t 1 α+E[ni|d i ] +m t otherwise
The Average Treatment effect on the Treated (ATT) measures the additional outcome change experienced by those who received the treatment compared to those who did not, as defined by Blundell and Costa Dias (2009).
= [y¯ 1,t 1 −y¯ 1,t 0 ]−[y¯ 0,t 1 −y¯ 0,t 0 ] where ¯y d,t is the average outcome over groupdat timet.
The Difference-In-Difference (DID) estimator, as detailed by Angrist and Pischke (2009), can be empirically analyzed in scenarios where treatment occurs at a single point in time, involving two periods (before and after treatment) and two groups (treated and control) The DID method is represented by the equation: yit = β0 + β1ds + β2dt + δ(ds∗dt) + εit, where yit denotes the outcome for individual i at time t The variable ds indicates potential pre-policy differences between the treated and control groups, with ds being a dummy variable (1 for treated, 0 for control) The variable dt accounts for aggregate factors affecting outcome changes over time, with dt also being a dummy (1 for the policy year or later, 0 for before the policy) The key coefficient of interest, δ, measures the difference in changes between the treated and control groups before and after the policy implementation.
Applying this method in our analysis, we use panel data from 2005 to
In 2012, domestic firms in the treated group benefited from a tax reduction from 28% to 25%, while the control group, consisting of foreign firms and other domestic firms not eligible for the tax reduction, remained unaffected The analysis employs a standard Difference-in-Differences (DID) approach, examining two groups over an eight-year period from 2005 to 2012 This methodology incorporates a comprehensive set of time dummies and firm-specific effects to enhance the model's accuracy.
The Difference-in-Differences (DID) regression model can be enhanced by incorporating individual-specific effects within a panel data framework This allows for a more nuanced analysis over T time periods, represented by the equation yit = ci +.
The equation ∑ t=1 γtdt + δTreatit + Xit + εit (1.2) illustrates the model's structure, where cicaptures individual-specific effects and dt represents time dummies The variable Treatit is defined as the product of ds and dt, equaling 1 when the individual is treated (ds=1) and during or after the treatment period (dt=1); otherwise, it is 0 Additionally, Xit consists of a vector of explanatory variables that may influence the outcome.
The model can be enhanced by incorporating a firm-specific time trend (git), transforming it into a Random Trend Model According to Wooldridge (2002), this approach permits each individual to possess a unique time trend, which must be integrated into the model Consequently, both ci and gi may exhibit correlation with the explanatory variables, represented by the equation yit = ci + git +
To address a Random trend model, Wooldridge (2002) suggests employing the First Differencing method to remove firm-specific effects This approach simplifies the equation to the standard unobserved effects model, where the difference between current and previous values yields a consistent term.
With this model, we can apply fixed effects or first-differencing meth- ods to estimateδ.
The Difference-in-Differences (DID) approach relies on the assumption that both treated and control groups share common trends or are affected by the same macroeconomic shocks; if this assumption is violated, the estimated treatment effect may be inconsistent (Blundell and Costa Dias, 2009) While there is no standardized method to test this common trend assumption, graphical plots can help identify potential differential trends, and the Granger causality test can assess the presence of anticipation effects (Wing et al., 2018) To investigate anticipation effects, leading values of the treatment variable can be incorporated into the model.
The equation ∑ m=1 βmTreat i,(t+m) +Xit+εit (1.5) estimates the lead treatment effect for m periods prior to the policy year The underlying hypothesis suggests that future policy changes should not influence current outcomes If the coefficient βm is statistically significant, it indicates the presence of an anticipation effect, which may challenge the validity of the assumption and render the model problematic.
Testing time-varying treatment effects can be effectively conducted using Difference-in-Differences (DID) regression Researchers can enhance their analysis by incorporating lagged treatment variables into the standard DID model, allowing them to explore how the treatment effect changes over time following its implementation.
The treatment effect δ measures the immediate impact of tax cut policies, while the lagged treatment assesses additional effects that may arise in subsequent periods A positive immediate effect paired with a negative lagged value suggests that the initial benefits of the policy diminish over time Conversely, if both the immediate effect and lagged value are positive, it indicates that the initial benefits of the policy are amplified as time progresses (Wing et al., 2018).
In our empirical analysis, we employ a Difference-in-Differences (DID) model using fixed effects panel data regression This approach allows us to account for both firm-specific time trends and individual firm effects by applying fixed effects estimation to first-differenced variables By first-differencing, we effectively eliminate the time trend associated with each firm, while fixed effects estimation addresses the issue of firm-specific effects within the panel data framework.
Data and variable measurement
Data
The data for this study was sourced from the Annual Firm Survey conducted by the Vietnamese General Statistical Office (GSO), which began in 2000 and has produced seventeen years of data Initially, the survey encompassed all firms in the economy, but due to the rapid increase in the number of firms by 2005, the GSO adopted a mixed approach of population and sample surveys The population survey includes state, foreign, and non-state firms with a workforce above a variable threshold set annually, while specific economic activities and provinces with fewer firms are fully surveyed Non-state firms, on the other hand, are selected through a random sampling process that ranges from 10% to 50% of the total, depending on labor size and location.
The data set includes three key components: first, details for firm identification such as name, address, business type, and economic activities; second, information regarding labor and labor income; and third, insights into asset and business performance, encompassing assets, taxes, investments, R&D expenditures, revenue from goods and services categorized by economic activities, as well as costs and profits.
In this study, we analyze a dataset that has grown significantly from 42,307 observations in 2000 to 455,238 in 2015 Focusing on the period from 2005 to 2012, we compile annual survey data, with each firm's unique tax code facilitating the matching process across years due to the absence of firm codes in some instances Our analysis reveals a consistent presence of 17,131 firms each year during this timeframe, culminating in a comprehensive panel dataset spanning 8 years and comprising 137,048 observations.
Variable measurement
Our variable measurements primarily rely on the literature and the Annual Firm Survey conducted by GSO, as detailed in section 1.4 We have established four dependent variables, with capital stock defined as the physical assets present at a specific point in time.
Capital stock refers to a firm's plant, equipment, and other fixed assets, as defined by Rudolf and Zurlinden (2008) In this study, capital stock is measured as the total value of fixed assets at the end of each year To effectively analyze the linear relationship between independent and dependent variables, we utilize logarithmic transformation for the firm's capital stock.
Research on the investment rate has been extensive, defined as the ratio of the change in net fixed assets over a year to the net fixed assets at the year's start (Kadapakkam et al., 1998; Dwenger, 2009) It can also be described as net investment from new long-term investments minus depreciation, divided by the average total assets (Silva et al., 2012) Zhang et al (2018) adopted this definition while adjusting for the price index of fixed investment to the year 2000, allowing for the conversion of net asset values to a constant price Some studies utilize both the natural logarithm of investment levels and the investment rate to analyze firm behavior (Zwick and Mahon, 2017) This paper calculates the investment rate by assessing the annual change in fixed assets, scaled by lagged fixed assets, consistent with previous research.
The employment growth rate is determined by the change in total labor compared to the number of workers at the beginning of the year Labor income reflects the average income per worker that companies provide, representing the average wage in the labor market Wage growth is calculated by assessing the annual change in wages from the current year compared to the previous year's labor wages.
In our analysis, we introduce two control variables: profitability, defined as the ratio of pre-tax profit to total revenue, and capital intensity, calculated as total assets divided by total labor These variables influence a firm's investment behavior, capital stock, employment, and labor income payments, while remaining unaffected by tax policy assignments Therefore, they are essential for inclusion in our Difference-in-Differences (DID) regression model.
The impact of tax cut policies can differ based on a firm's characteristics, such as size and liquidity ratio, as discussed in section 1.7 In Vietnam, firms are classified as small-sized if their total assets are 20 billion VND or less, or if they employ 200 workers or fewer in the agricultural, manufacturing, and construction sectors For the commercial sector, the thresholds are total assets of 10 billion VND or less, or 50 employees or fewer In our study, we use total labor as the classification criterion and define a dummy variable, SMALL, which equals 1 for small and medium-sized enterprises and 0 for larger firms.
A short description of the variables is given in table (1.1) below.
4 Small and medium-sized firms indicators were established through Decree no.56-2009-ND-CP.
Table 1.1: Variables used in empirical regression
Capital stock Total fixed asset end of each year.
Investment rate The annual change in fixed asset, scaled by lagged fixed asset
Employment rate The change in total labor divided by the number of labor beginning the year
Wage growth Annual change of wage in current year to lagged labor wage.
Capital intensity equals to total asset divided to total labor.
Profitability the ratio of pre-tax profit and total revenue
SMALL as a dummy variable which equals to 1 if it is small firms and 0 otherwise
CF equal to 1 if in the year 2008, firms have low cash flow rate and equal to 0 otherwise.
The liquidity ratio is assessed using the cash flow rate, which is calculated by adding pre-tax profit to depreciation and normalizing it by the lagged fixed asset Normalizing the liquidity ratio facilitates comparisons among firms of different sizes Additionally, a dummy variable, cash flowCF, is established, assigning a value of 1 to firms with a cash flow rate below the median in 2008, while firms above the median receive a value of 0.
The table (1.2) below shows the description of variables in different groups.
Treatment and control groups
In 2009, a tax cut policy reduced the corporate tax rate for domestic firms from 28% to 25% However, this policy did not extend to domestic firms already benefiting from a preferential tax rate of 20% or lower.
Variables Treated firms Non-treated firms
Mean Std Dev Mean Std Dev
Newly established firms in challenging socio-economic regions, as well as those entering high-tech and scientific research subsectors, benefit from a reduced tax rate of 10% for fifteen years Additionally, firms with investment projects in these difficult areas enjoy a 20% tax rate for ten years Agricultural service cooperatives also qualify for the 20% tax rate, promoting growth and development in these sectors.
The study categorizes firms into two groups: the treatment group, which includes domestic firms without preferential tax benefits, and the control group, comprising foreign firms and domestic firms that do receive these tax advantages The dataset allows for the differentiation of domestic and foreign firms through ownership variables, while the classification of domestic firms with or without preferential tax is based on their addresses, economic activities, and tax payments for the year Table 1.3 presents the treatment and control groups related to the 2009 tax cut policy based on this information.
5 The list of extremely difficult and difficult socio-economic condition areas in Vietnam was deter- mined in the Decision number 190/2005/QD-TTg in July 2005.
Table 1.3: Tax cut treatment and control group
Groups Descriptions Number of firms
Treatment group Domestic firms without preferential tax
Domestic firms with preferential tax
Foreign firms +Domestic firms with preferential tax
Data analysis and results
Capital stock and Investment rate
Figure 1.2 illustrates the average capital stock and investment rates for both treated and control groups, as defined in Table 1.3 The treated firms consist of domestic entities without preferential tax benefits, while the control group includes foreign firms and other domestic companies Both groups exhibit a similar upward trend in capital stock, although treated firms have a lower capital stock compared to control firms Notably, since 2009, the capital stock gap has been narrowing, with treated firms showing a steeper growth trajectory Additionally, the investment rate for treated firms has consistently surpassed that of control firms, with a significant disparity observed in the policy year of 2009.
In 2009, during the global economic crisis, both domestic and foreign firms experienced a decline in investment rates; however, the impact was less severe for domestic companies due to government tax rate cuts, which helped mitigate the challenging business environment.
Table 1.4 presents alternative specification tests for the Difference-In-Difference method, analyzing two dependent variables: the log of capital stock and the investment rate The first column displays the baseline estimates, which incorporate firm-specific and time effects using a fixed effect model with robust options The results indicate that the 2009 tax reduction policy significantly impacted firms' total capital stock, yielding an estimate of 0.414 log points, equivalent to a 51.28 percent increase In contrast, the treatment effect on the investment rate is found to be insignificant.
In our analysis, we incorporate a firm-specific time trend into the model to capture the varying evolution of capital stock and investment rates across different firms This addition is crucial as it distinguishes the behaviors of domestic firms, which are treated, from the predominantly foreign control firms The presence of unobservable characteristics may also contribute to the differing trends in these dependent variables Therefore, we adopt model 1.3 as our benchmark model to reflect these trends accurately.
The analysis reveals that capital stock in domestic firms increased by about 0.119 log points, translating to a 12.6% growth, which is significantly higher than that of foreign firms Additionally, when accounting for firm-specific time trends, the investment rate for domestic firms shows a substantial increase of 21.6% prior to the tax reform, indicating a markedly positive investment environment compared to their foreign counterparts.
6 Converting the log points to percentage points by the formula:%4y = exp( β ˆ ) − 1
Table 1.4: Capital stock and investment rate response to the corporate in- come tax change
Dependent variable: Log capital stock
Firm specific effect Yes Yes Yes
Time effect Yes Yes Yes
Firm specific time trend No Yes Yes
The analysis presented in Column (1) includes estimates from a Difference-in-Difference regression that accounts for both firm-specific effects and time effects In Column (2), a firm-specific time trend is incorporated to better adjust for disparities between domestic and foreign firms that may influence the evolution of capital stock and investment over time Column (3) further enhances the model by introducing additional control variables to refine the results.
Robust standard errors in parentheses.
The third column shows the estimates with the same model as column
After accounting for key factors such as a firm's profitability and capital intensity, domestic firms show a significant post-policy increase in capital stock and investment rates, with a rise of 0.092 log points (equivalent to 9.63%) and a 19.3% increase, respectively, when compared to foreign firms.
The implementation of the tax reduction policy amidst an economic crisis raises concerns, particularly regarding the differing behaviors of domestic and foreign firms due to ownership structures To address this issue and ensure robust findings, we conducted an additional regression analysis excluding foreign firms from the ineligible group while maintaining the same eligible group observations Despite the significant disparity in the number of non-treated domestic firms (3,472) compared to treated domestic firms (121,848), both groups exhibit similar characteristics, indicating no significant differences in ownership among them.
According to the benchmark model (equation 1.3), the estimates for capital stock and investment rate are 0.062 log points (equivalent to 6.39%) and 27.6%, respectively, as illustrated in Table 1.5 These findings reinforce our earlier conclusions that tax policy positively impacts these dependent variables, regardless of the varying ownership structures of firms and the prevailing economic crisis.
Table 1.5: Capital stock and investment rate response to the corporate income tax change with no foreign firms in the control group
Log of capital stock Investment rate Treatit
Firm specific effect Yes Yes
Firm specific time trend Yes Yes
The treated group and control group are domestic firms which are homogeneous in terms of ownership.
Robust standard errors in parentheses.
In section 1.3, we outline our approach to verifying the parallel trend assumption in the Difference-in-Differences (DID) method, which ensures that the differences between the treated and control groups remain consistent over time in the absence of treatment Our graphical analysis partially supports this common trend assumption, as illustrated by the nearly parallel plots of capital stock and investment rates (see Figure 1.2) Additionally, we assess the anticipation effects of the tax cut policy; if these effects are negligible, the validity of our DID regression is affirmed.
Table 1.6: Lead and lagged treatment effect of 3% tax reduction policy in 2009
Notes: Column (1), column (2) and (3) we regressed based on the equation 1.5 and 1.6 Control variables as profitability and capital intensity are included in the model.
Robust standard errors in parentheses Significance level: *% **=5%.
Table 1.6 presents specification tests to evaluate the lead of the treatment effect and assess additional impacts in the years following the policy's conclusion We conduct an anticipation effect test for one period using equation 1.5, incorporating control variables The results in Column (1) indicate that the estimates for the anticipation test are insignificant for both capital stock and investment rate, confirming the validity of the underlying assumption.
Time-varying treatment effects are also shown in column (2) and (3).
The impact of tax cut policies on investment rates has diminished over two consecutive years, particularly evident in 2009, coinciding with the global economic crisis that generally led to a decline in investment rates Our findings suggest that the tax cut policy in 2009 helped prevent a significant drop in domestic firms' investments compared to their foreign counterparts While foreign firms opted to delay their investments during the recovery period, domestic firms acted promptly to capitalize on the tax cut opportunity Consequently, the initial gap in investment rates between domestic and foreign firms widened, only to narrow in subsequent years.
Employment and labor wage
We apply the same method for employment and labor income analysis.
The benchmark DID model and robustness testing model are utilized in Table 1.7 to analyze employment rates and labor income, using the same database and time period Results for employment growth and labor income growth are presented in Columns (1) and (4), incorporating firm-specific effects, time effects, and firm-specific time trends along with control variables Additionally, Columns (2) provide further insights into the treatment effects.
(5) represent the anticipation effect Column (3) and (6) show the time-varying treatment effect for the two years after the treatment.
The disparity in employment growth between domestic and foreign firms remained unaffected by the 2009 tax reduction Nonetheless, a persistent negative gap between these groups emerged in the subsequent two periods following the tax cut Our findings in Vietnam contrast with the results reported by Shuai and Chmura.
A study conducted in 2013 revealed that reducing corporate income tax can lead to immediate job creation within the same year, with no significant effects observed in subsequent years.
Table 1.7: Treatment effect on employment and labor income
Employment growth Labor income growth
Notes: Column (1), column (2) and (3) we regressed based on the equation 1.5 and 1.6 Control variables as profitability and capital intensity are included in the model.
Robust standard errors in parentheses Significance level: *
The corporate tax cut has significantly influenced employee compensation, leading domestic firms to pay approximately 17% more than their foreign counterparts This tax incentive has resulted in a higher investment rate for domestic firms, which in turn has contributed to greater wage growth compared to foreign firms However, the long-term effects of the tax cut on labor wages show no significant estimates beyond the initial policy year These findings align with earlier observations regarding the investment rate, indicating that while firms increased both investment and wages following the 2009 tax reduction, the impact on investment diminishes over time, resulting in non-significant lagged estimates of labor income.
Ljungqvist and Smolyansky (2018) employed the Difference-in-Differences (DID) method to examine the impact of tax increases on employment and income in the US Their findings revealed that tax increases initially resulted in a decline in both employment and total wages However, in subsequent years, employment and income growth rates became statistically similar between the affected and control firms This conclusion aligns with our previous results.
Extensions
Our analysis differentiates treatment effects based on firm characteristics, focusing on factors such as firm size and liquidity ratio Building on previous research that highlights varying impacts due to firm-level attributes, we also evaluate the policy effects across different economic sectors.
Table 1.8 illustrates the varying treatment effects across different dependent variables Utilizing the benchmark Difference-in-Differences (DID) model from equation 1.3, we incorporate additional interaction terms that account for the treatment effect in relation to the firm's size and liquidity ratio.
Table 1.8: Treatment effect distinction of tax cut policy in 2009
Dependent variables Treat it Treat it ×
The results presented are derived from the DID regression analysis outlined in equation 1.3, incorporating an interaction term that examines the treatment effect alongside firm size and liquidity ratio Each regression accounts for firm fixed effects, time effects, and firm-specific time trends, along with various explanatory variables.
Robust standard errors in parentheses Significance level: *% **=5%.
In our analysis, we introduce an interaction term that combines the treatment effect with the size of the firm Our dataset reveals that out of 137,048 firms, 79,008 are classified as small, with 75,008 of these belonging to the treated group The classification of firm size and the dummy variable SMALL are defined in section 1.4.2.
Table 1.8 indicates that small firms exhibit a significant positive response to investment incentives, with an 18.7% increase in investment rates and a 3.3% rise in total capital stock Due to their limited assets and financial constraints, small firms face challenges in accessing loans (Egger et al., 2018) Therefore, corporate income tax credits can play a crucial role in stimulating their investment activities This aligns with Zwick and Mahon (2017), who found that temporary tax incentives have a substantially greater impact on small firms compared to larger ones Additionally, our results support the findings of Cerda and Larrain (2010), while contradicting those of Egger et al (2018) as discussed in the literature review.
Recent data indicates that employment growth shows no significant influence from policies targeting firm size However, small and medium-sized enterprises (SMEs) offer wages that are approximately 14% higher than those of larger firms This trend suggests that SMEs are more responsive to tax cuts in investment rates, leading to the expectation that they will provide higher compensation to their employees.
The second extension of our analysis focuses on the liquidity ratio, utilizing cash flow (CF) as a proxy, as discussed in section 1.4.2 The expected sign of this estimate remains uncertain; however, cash-constrained firms might leverage tax reductions to invest, thereby enhancing their cash flow through lower tax payments, which could result in a positive estimate Supporting this notion, Zwick and Mahon (2017) provide evidence that low-liquidity firms experience a significantly larger impact compared to their high-liquidity counterparts.
Cash-constrained firms are generally less inclined to invest compared to those with high cash flow rates, primarily due to financial limitations Firms with robust cash flow readily seize investment opportunities, which can lead to a negative overall investment estimate for their less-capitalized counterparts As shown in Table 1.8, the interaction terms in rows (2) and (4) are statistically significant and negative, indicating that firms with low cash flow ratios experience a markedly smaller investment impact This finding aligns closely with Zhang et al (2018), who utilized the same Difference-in-Difference method and discovered that China's value-added tax reform had a more pronounced investment effect on firms with above-median cash flow ratios.
Our analysis reveals that changes in employment growth and labor income growth do not vary significantly with the firm's liquidity ratio, as indicated by the lack of significance in the interaction terms with this variable.
Table 1.9: Treatment effect of tax cut policy by economic sectors
indicates a significance level of 5%.
This analysis examines the intensity of treatment effects across various economic sectors, categorizing them into three groups: Agriculture (including Forestry and Fishing), Manufacturing (encompassing Manufacturing and Construction), and Services We apply the standard Difference-in-Differences (DID) model to each industry sector independently for every outcome variable.
Domestic firms in the agriculture sector are not benefiting from tax policy, while the manufacturing sector experiences increased capital stock and investment rates Notably, the service sector shows a significant response to tax policy changes, with treatment effects on capital stock and investment rates surpassing those in manufacturing A corporate tax cut resulted in a remarkable investment rate surge to 44.5% in the service sector, four times that of manufacturing, and a 15% increase in capital stock, double that of manufacturing This finding contradicts Djankov et al (2010), who reported that corporate tax rates are linked to investment in the manufacturing sector but do not affect the services sector, where informal activities are more common.
The impact of the tax cut on employment growth shows significant effects solely within the service sector, while its influence on income growth is notably significant only in the manufacturing sector.
Conclusions
In 2009, the Vietnamese Government implemented a significant reform to its corporate income tax policy, reducing the rate from 28% to 25% for domestic firms, aligning it with the previously privileged rate for foreign companies This policy aimed to enhance competition among firms and stimulate economic growth To assess the impact of this reform on capital stock, investment rates, employment rates, and labor income, we employed a Difference-In-Difference approach, analyzing a panel of firms over an eight-year period from 2005 to 2012 The results indicate that the reform served as a strong incentive for increased economic activity.
The classification of industries in Vietnam is guided by Decision number 10/2007/QD-TTg, issued by the Prime Minister on January 23, 2007 This framework aims to enhance investment and capital stock, ultimately boosting labor income for domestic firms in comparison to foreign competitors Consequently, the tax reduction implemented in 2009 is recognized as an effective policy that helped mitigate the decline in investment rates among domestic firms during economic downturns.
Hall and Jorgenson (1967) highlight that the cost of capital encompasses both the discount rate and the depreciation rate A corporate tax credit can effectively decrease the depreciation base by the amount of the tax credit, resulting in a lower rental value of capital inputs This dynamic can stimulate an increase in capital stock and investment Evidence supporting this theory is observable in Vietnam, particularly in the context of its recent tax cut policy.
2009 led to 9.2% and 19.3% higher in capital stock and investment rate respectively.
Sandmo (1974) highlighted the impact of corporate income tax on capital stock composition, noting that such taxes can distort pretax factor prices, leading to a potential reduction in rental rates compared to wage rates, which encourages firms to adopt a more capital-intensive approach Our findings align with this perspective, revealing an increase in capital stock without a corresponding rise in employment, indicating that Vietnamese firms are prioritizing capital investment over labor in response to tax cuts This trend supports the Vietnamese government's objective of enhancing capital intensity as part of its modernization goals within tax reform strategies.
The service sector experiences the greatest benefits from this policy compared to manufacturing, while the agriculture sector shows no significant impact Additionally, small-sized firms with high liquidity ratios demonstrate a stronger responsiveness to this reform.
Figure 1.2: Mean of capital stock and investment rate by groups
2004 2006 2008 2010 2012 year treated firms control firms
2004 2006 2008 2010 2012 year treated firms control firms
The treated firms include domestic firms without preferential tax, the control firms include foreign firms and other domestic firms with preferential tax.
Participation and the role of sunk cost: An analysis at firm-level data in Vietnam
Introduction
Since the implementation of Economic Reforms in 1986, Vietnam has transitioned from a centrally-planned economy to a socialist-oriented market economy, prioritizing trade liberalization and economic integration in its strategies The government focuses on an export-oriented approach and attracting foreign investment as key pillars of its policies These reforms have opened up opportunities in international trade, driving economic development and technological innovation From 1995 to 2010, Vietnam's export turnover surged by an average of 21% annually, reaching 72.2 billion USD in 2010, which accounted for nearly 71% of the country's GDP.
Three significant milestones have greatly influenced Vietnam's export performance: the signing of the Vietnam-US Bilateral Trade Agreement (BTA) in December 2001, which opened the world's largest market to Vietnamese exporters; the implementation of the amended Commercial Law in August 2001, allowing all legal entities to export most goods without the need for a license or quota; and Vietnam's accession to the WTO in November 2006, officially becoming a member in January 2007, which further integrated the country into the global trade system.
Vietnam's accession to the WTO in 2007 marked a significant step towards deeper integration into the global economy, leading to a surge in export opportunities This membership resulted in an export boom for the country during the years 2007 and 2008, enhancing its economic growth and global trade presence.
Vietnam's export structure is primarily composed of primary and labor-intensive products, particularly in the agriculture, fishery, forestry, and footwear sectors The growing presence of foreign direct investment (FDI) firms in the export market intensifies competition for domestic companies Additionally, Vietnam's involvement in the assembly and simple processing stages of the international value chain limits its ability to generate high value-added income from exports.
The Vietnamese Government has implemented various export promotion policies to encourage participation in the export sector, particularly within industry and high technology These initiatives include reductions in corporate income tax and export tax, as well as prioritizing land use for exporting firms By 2017, nearly 24,000 firms were engaged in exporting, reflecting an annual growth rate of 7.1% from 2012 to 2017 The government continues to prioritize export participation across both state-owned and small to medium-sized non-state businesses Key support measures include enhancing infrastructure for industrial parks and export processing zones, offering tax incentives and exemptions for firms that import goods for export production, and providing credit programs for exporting companies.
2 Vietnamese Custom website: https://haiquanonline.com.vn.
3 The Decree of 108/2006/ND-CP in 2006 for guiding the Investment Law.
4 The Decree of 134/2016/ND-CP in 2016 for implementing the law on export and import tax.
5 The Decree of 75/2011/ND-CP in 2011 for Investment credit and export credit.
Numerous empirical studies have examined the decision-making process behind a firm's participation in export markets, highlighting the significance of entry sunk costs as a critical factor influencing this choice (Roberts and Tybout, 1997; Bernard and Wagner).
Research indicates that more productive firms are more likely to engage in exporting, as they can better absorb the sunk costs associated with entering foreign markets In contrast, less productive firms often refrain from exporting due to concerns that their profits may not sufficiently offset these additional expenses.
This paper investigates the impact of sunk export costs on the export participation of Vietnamese manufacturing firms Sunk costs associated with exporting, such as identifying foreign buyers, understanding market regulations, and establishing transportation channels, can pose significant barriers for these firms To analyze this relationship, we employ a dynamic Probit model that incorporates previous export status and utilizes Wooldridge's methodology to address the initial condition problem Our findings align with existing literature, confirming that substantial sunk costs are indeed a critical factor in export entry decisions.
In Vietnam, firms exhibit diverse characteristics that influence their likelihood of becoming exporters, even within the same macroeconomic context Key factors such as productivity, age, wage levels, and capital intensity play crucial roles in export decisions Our analysis reveals that firms with higher productivity and greater capital intensity are more inclined to engage in export activities, whereas smaller firms tend to have a lower probability of participating in exports.
The impact of spillover effects on a firm's export participation is analyzed through changes in sunk and fixed costs We investigate the influence of foreign direct investment (FDI) firms within the same province and in other provinces, adjusted for distance and sub-sector similarities Drawing on the research of Maurseth and Medin (2017), we highlight that spillovers can alter both sunk and fixed export costs They argue that fixed costs, such as those associated with customs declarations, can decrease as firms gain experience, affecting not only new entrants but also established exporters To explore this further, we differentiate the spillover effects for various firm types by incorporating interactions between spillover variables and firm classifications.
This paper is structured into several key sections: the literature review on exporting participation is discussed in the next section, followed by a general econometric approach in Section 3 Section 4 provides a detailed overview of data and variable measurement, while Section 5 presents the results Finally, conclusions are drawn in Section 6.
Literature Review
Numerous studies have explored firm-level engagement in international trade, focusing on the differences between exporters and non-exporters A significant contribution to this field is the work of Roberts and Tybout (1997), which presents a dynamic discrete choice model analyzing the impact of entry sunk costs on export decisions Their research, based on empirical data from Colombia, reveals that firms face considerable sunk costs when deciding to export, with prior exporting status increasing the likelihood of future exports by nearly 60 percentage points Additional studies, such as those by Bernard and Wagner (2001) and Arnold and Hussinger (2005) in Germany, as well as Bernard and Jensen (2004b) in the United States and Kandilov and Zheng (2011) in Ireland, have further examined the role of sunk entry costs in export decision-making across various countries.
Research highlights the significance of firm-level heterogeneity in export decision-making, with productivity variations among firms influencing their exporting behaviors Melitz (2003) employs a dynamic industry model to examine how international trade facilitates the reallocation of diverse firms within an industry.
In this model, the author highlights that more productive firms with lower marginal costs can effectively cover entry fixed costs, making them more likely to become exporters Additionally, reduced exporting costs lower the productivity threshold required for firms to engage in export activities Melitz and Ottaviano (2008) further analyze the interplay between market size, trade, and productivity within a monopolistically competitive framework They argue that increased market size and trade intensity enhance competition, influencing the selection of producers and exporters based on their heterogeneous traits Their findings indicate that larger markets foster intense competition, resulting in lower average mark-ups and increased productivity.
Subsequent empirical studies reinforce the notion that more productive firms tend to self-select into foreign markets This correlation is evidenced by research from Arnold and Hussinger (2005), which employs a dynamic Probit model to analyze exporting in Germany, as well as studies by Tuano et al (2014) in the Philippines and Amornkitvikai et al.
In 2012, a study highlighted the correlation between productivity and export participation in Thailand Conversely, research by Greenaway and Kneller (2004) found negligible effects of productivity on export participation in the United Kingdom, a finding echoed by Clerides et al.
In examining the productivity advantages of exporting compared to non-exporting firms, the concept of learning-by-exporting emerges as a key factor This theory suggests that firms experience productivity increases after they begin exporting, rather than prior, due to the technical expertise gained from foreign buyers and the intense competition present in export markets Research supporting this hypothesis includes studies by Clerides et al (1998), Arnold and Hussinger (2005), Blalock and Gertler (2004), and Bernard and Jensen (2004b).
Research indicates a correlation between workforce skill levels, managerial characteristics, and a firm's export behavior Amornkitvikai et al (2012) demonstrate that skilled workforces enhance export participation among Thailand's manufacturing SMEs Similarly, Obben and Magagula (2003) utilize a Logit model in Swaziland, revealing that older managers are more likely to lead exporting firms, albeit at a diminishing rate The study also shows that managers who perceive greater risks in export sales compared to domestic sales are more inclined to engage in exporting, suggesting that managers in exporting firms tend to be more risk-tolerant than their non-exporting counterparts.
A study conducted in 2012 analyzed the factors influencing the propensity to export among textile firms in the UK and Portugal, utilizing a dataset from both countries The research employed methods such as analysis of variance (ANOVA), multiple regression analysis, and principal component analysis Findings revealed that in Portugal, the education level of managers significantly impacts export propensity, while in the UK, the age of managers and their perception of costs are the primary influencing factors.
Capital intensity significantly influences a firm's export behavior, as suggested by Sinani and Hobdari (2010), who argue that it reflects the technological gap between exporters and non-exporters Consequently, firms with higher capital intensity are expected to produce superior quality products and are more inclined to enter international markets This hypothesis is supported by empirical findings from Estonian firms in Sinani and Hobdari's study Similarly, Wagner (2011) observed comparable results in the analysis of German manufacturing industries, reinforcing the link between capital intensity and export performance.
Studies have yielded mixed results on the relationship between capital intensity and export propensity While some research, such as (2019) and Soderbom (2000), found a positive and significant correlation in Ghana, others, including Srinivasan and Archana (2011) and Gourlay et al (2005), discovered a significantly negative link in India and the UK, respectively The latter findings are consistent with trade theory, suggesting that India's comparative advantage lies in labor-intensive industries, whereas the UK's comparative advantage is in less capital-intensive services.
Another heterogeneous firm characteristic involved in export probabil- ity is the firm’s size Papers have overwhelmingly found that larger firms (which is measured in terms of the number of labor) are likely to become exporters (Muuls and Pisu (2009) for Belgium case, Arnold and Hussinger (2005) for German case,Roberts and Tybout (1997) for Colombia case) Moreover, positive estimates of firm age and capital stock in the export decision are recorded Roberts and Tybout (1997) explained that capital stock may represent productivity and firm age represents the cost differences among producers Older firms tend to be more competitive in the world market due to the advantage in cost and thus, more likely to perceive larger return when engaging the foreign market.
Not only the heterogeneity of firms is the cause of export probability, but also other factors at region level or sector level can change the firm export deci- sion, such as the export price relative to domestic output price to capture the foreign market attraction (Roberts and Tybout (1997)) or some externalities including the activities, locations of other firms can decrease the cost as firms access to foreign markets and induce export participation Such externalities are the export activity in the same state but different industries or export activities in the same industry and state (Bernard and Jensen (2004b)) The work of Maurseth and Medin (2017) focuses on the export participation that accounts for the sunk export cost for a par- ticular country or a particular product This work also examines the spillover effects of the number of other firms producing the same product or other exporting firms with different products but the same destination The prominent point of Maurseth and Medin (2017) is that the authors try to analyze these spillover effects which distinguish the effect between the entrants and continuing exporters The idea is that not only the export sunk cost but also fixed costs are affected Hence, the en- trants whose exporting participation are influenced by both sunk cost and fixed cost would behave differently from the continuing exporters which only have the fixed cost Employing the Norwegian firm-level panel data in the seafood sector, the re- sults support the hypothesis of market-specific sunk costs and spillover effects are exposed to both the entrants and continuing exporters.
Moreover, the spillover effects on export participation is also men- tioned in Buck et al (2007) This paper uses the Heckman sampling model to analyze the export spillover effect from multinational firms in China for the period 1998-2001 The result shows that multinational firms create positive export par- ticipation for Chinese domestic firms through a wide range of spillover channels including labor mobility, spatial agglomeration and technological imitation Ngo
(2019) examines the factors affecting the export spillover effect of FDI in three years, from 2016 to 2018, in Vietnam The paper confirms that foreign firms have positive effect on export propensity in domestic firms However, this spillover effect is different depending on different characteristics of domestic firms.
Research on the determinants of export propensity and performance in Vietnam has been conducted, focusing on various factors affecting export participation and performance at the firm level using different methodologies However, most studies overlook the role of sunk export costs Nguyen et al (2015) investigated the impact of innovation and productivity on export participation among Vietnamese SMEs, finding a positive correlation between the two using Probit methodology Similarly, Nguyen et al (2008) addressed the potential endogeneity of the innovation-export relationship using instrument variables in a binary Probit method, employing the number of employees with a college education and investment strategies as instrumental variables.
2005, results indicate that innovation is significantly positive to export propensity.