Caught in the Middle

What is the impact of the US–China trade war on Taiwanese direct investment (TDI) in mainland China, by Taiwanese business people, or “Taishang”?

Are the Taishang incentive policies issued by Beijing and implemented by the Chinese local governments effective in keeping Taishang from withdrawing from mainland China, especially during the trade war period?

With the newly available TDI data from the Republic of China Ministry of Economic Affairs, we utilized the difference-in-difference (DID) estimation upon the monthly data of TDI within 31 provinces in China in 2018.

We found that the third wave of the trade war did not have any statistically significant impact on cumulative TDI or monthly manufacturing TDI, despite some negative impact on the non-manufacturing TDI.

We also found that the Taishang incentive policies play a significant role in attracting Taishang and was not weakened by the outbreak of the trade war.

Introduction

What is the impact of the US–China trade war on Taishang? Are the Taishang incentive policies issued by Beijing and implemented by the Chinese local governments effective in keeping Taishang from withdrawing from mainland China, especially during the trade war period?

To better explore these questions, we first review the cross-Strait economic relationship and analyze the rationale on the impact of trade war and Taishang incentive policies on Taishangs’ movement in and out of mainland China.

None of the existing studies have empirically examined the effect of the coexistence of the trade war and the Taishang incentives policies.

Existing qualitative analyses have shared the concern that the close trade relationship across the Strait and its important role in global supply chain, and the volatile US–China relationship make it unavoidable for Taishang to suffer loss in this trade war.

Thus, Taishang are expected to be the major losers in the US–China trade tensions. Despite this, Taishang have been granted significant preferential policies by the Chinese national and local governments to help Taishang maintain sustainable development in mainland China.

Therefore, during the US–China trade war, Taishang faced both risks and opportunities in mainland China.

It is important to conduct serious empirical investigation on how the interaction between the risks (i.e., the trade war) and the opportunities (i.e., the Taishang-preferential policies) impact the Taishangs’ movement in and out of China.

To better explore these questions, we first review the cross-Strait economic relationship and analyze the rationale on the impact of trade war and Taishang incentive policies on Taishangs’ movement in and out of mainland China.

Then, we utilize the monthly TDI data and conduct empirical analysis on whether Taishangs’ movement in and out of China has been affected by the outbreak of the trade war, the issuing of the Taishang-preferential policies by the Chinese local governments, and the interactions between the two.

Next, we analyze our empirical findings. Last, but not the least, we summarize this article and make a conclusion.

Review of Cross-Strait Economic Relationship and its impact on Taishang

The cross-Strait economic relationship has been heavily influenced by the political and economic development on each side of the Strait, by the two countries’ bilateral relationship, and by global political and economic developments.

Since 1949, when the Chinese Communist Party (CCP) established the People’s Republic of China (PRC) and the Nationalist Party (KMT) established Taiwan under the name of the Republic of China (ROC), each side has claimed to be the sole legitimate government of China.

During the 1971 United Nations General Assembly, the international community switched its recognition of China from the ROC to the PRC.

Two decades later, the two sides of the Strait reached an agreement to disagree regarding to the sovereignty issue, which was exemplified by the 1992 Consensus across the Strait.

On top of these challenges, the Taishang faced a much more competitive market in mainland China, particularly given the rising competitiveness of Chinese domestic private firms and state-owned enterprises and the inflows of competitive foreign direct investment (FDI) from the more advanced economies.

According to this consensus, both sides agree that there is only “One China” and each side can have its own interpretation on what that nomenclature means.

Under the relatively peaceful environment across the Strait, the cross-Strait economic activities became increasingly active in the 1980s.

In the late 1980s, when the first group of Taishang moved out of Taiwan due to its rising labor cost and structural change, they arrived in mainland China to take advantage of the low wages, cheap land, and tax breaks.

This first group of Taishang greatly energized China’s economy under its marketization reforms.

The Taishangs’ development of mainland China from the 1990s and 2000s continued, and cross-Strait economic ties strengthened, although the legalization of the investment across the Strait did not happen until 2008.

Despite this, during this stage the Taishang faced many more challenges, due to the industrial upgrading made by mainland China, particularly under the pressure of the global financial crisis of 2008, China’s improved labor protection policies (e.g., the new Labor Contract Law), and the new environmental protection regulations.

On top of these challenges, the Taishang faced a much more competitive market in mainland China, particularly given the rising competitiveness of Chinese domestic private firms and state-owned enterprises and the inflows of competitive foreign direct investment (FDI) from the more advanced economies.

According to Chun-yi Lee, the Taishangs’ privileges at the early stage of Chinese economic openness has faded, and the Chinese private firm and state-owned enterprises increasingly are gaining more privileged treatments from Chinese local authorities.

Last, but not the least, since 2016, when the newly elected Taiwanese president, Tsai Ing-wen, took office, Beijing suspended official talks with Taiwan due to Tsai’s refusal to accept the 1992 consensus.

Since then, the cross-Strait relationship has been tense. This is another major challenge for the Taishang. Given the above major challenges faced by Taishang in recent years, has the trade war between the United States and China driven many Taishang to withdraw from mainland China?

Are Beijing’s Taishang-preferential policies helpful in keeping the Taishang from moving out of mainland China? Were Beijing’s Taishang-preferential policies still effective following the outbreak of the trade war?

The following three sections will address these questions respectively.

The Taishangs’ movement in and out of mainland China has important political implications for the cross-Strait relationship.

According to Robert F. Ash and Y. Y. Kueh, “economic integration is essentially a process of unification—the means whereby coherence is imposed upon previously separate, even disparate, geographical regions.”

As to the Taishangs’ political role, there has been a mainstream argument that Taishang serve as an important constituency in assisting with communications across the Taiwan Strait, despite the some arguments that Taishang are business animals and cannot be depended upon for political purposes.

Given the strategic importance of the Taishang, their movement in and out of China deserves closer investigation, particularly in light of the recent US–China trade war and the new Taishang incentive policies provided by the mainland Chinese national and local governments.

The Impact of the Trade War and Taishang Incentive Policies

Most observers suggest the trade war between the United States and China started in early 2018, although there had been a series of tariff increases on certain products before that.

With many rounds of trade talks between Washington and Beijing and claims of victory from both side, there are few signs that this trade war has truly run its course.

Despite the unpredictability of the trade war, the occurrence of the imbroglio and the PRC’s issuing of the Taishang-preferential policies in 2018 provide us an ideal environment to investigate (1) whether trade war caused the reduction of TDI in China; and (2) whether the Taishang incentive policies play a positive role in attracting TDI; and (3) whether the impact of the trade war weakened Taishang incentive policies.

Generally, existing studies divided the trade war up to now into three rounds. The first round began on 6 July 2018, when the United States imposed a 25-percent tariff on 34 billion USD of Chinese goods.

The second round began on 7 August 2018, when Washington released a revised version of tariffs on a final list of 16 billion USD–worth of imports from China and planned to implement a 25-percent tariff on them on 23 August.

The third round started on 17 September 2018, when Pres. Donald Trump announced a 10-percent tariff on 200 billion USD–worth of Chinese goods to begin on 24 September 2018 and aimed to increase to 25 percent by 1 January 2019.

On 2 December 2019, the United States and China agreed to a temporary truce and reached an agreement to refrain from increasing tariffs or imposing new tariffs for 90 days, up to 1 March 2019, while the two countries negotiated toward a larger trade deal.

While the first two rounds of the trade war did not impact Taishang significantly, the third wave greatly affected almost all export-oriented Taishang in China due to the tariff increase.

Despite the unpredictability of the trade war, the occurrence of the imbroglio and the PRC’s issuing of the Taishang-preferential policies in 2018 provide us an ideal environment to investigate (1) whether trade war caused the reduction of TDI in China; and (2) whether the Taishang incentive policies play a positive role in attracting TDI; and (3) whether the impact of the trade war weakened Taishang incentive policies.

Causal Mechanism of Trade War and TDI in Mainland China

In this article we divide Taishang into two categories: Taishang in the manufacturing sector and those in the nonmanufacturing sector.

We argue that Taishang in the manufacturing sector are negatively impacted by the trade war, although Taishang in other sectors (especially the service sector) may not be affected as much. During the past three decades, about 70 percent of Taishang in China operate in the manufacturing sector.

ROC–PRC–US trade mainly involves manufacturing production. This triangular trade relationship has been characterized as the following model: the United States delivers purchase orders to Taiwan firms located in Taiwan, which then request the Taishang (i.e., Taiwan-invested companies) in mainland China to produce the products ordered by the US market given the lower cost of producing in mainland China than in Taiwan.

As the Taishang in mainland China finish producing the products, they will export these products to the United States directly from mainland China.

Meanwhile, to produce the products to satisfy the needs of the US markets, Taishang in mainland China need to import large amounts of intermediate goods/components/parts from Taiwan.

This contributed to Taiwan’s high trade balance toward mainland China. A majority of Taishang in mainland China are mainly producing for the US market and serve as the original design manufacturers (ODM) for US manufacturing companies.

Therefore, Taishang in the manufacturing sector play an important role in mainland China’s exports to the US market.

Given the above trade model, when the trade war occurred, the exporting of the final products from mainland China to the United States would decrease due to the rise of tariffs.

The rising cost would lead to increased prices for the final products sold on the US market. Due to the supply-and-demand equilibrium, the rising prices would lead to the decrease of demands of such products imported from China and, thus, reduce the number of orders delivered by the US market to the Taiwanese companies in Taiwan.

In comparison, since the Taishang in the non-manufacturing sector are less tied to the ROC–PRC–US trade triangle, it is expected they would not be impacted much by the trade war.

This would reduce the number of orders that are forwarded to the Taishang in mainland China. As the value-added earned in the assembling factories in mainland China are usually small, the major way for them to make profits is to export large quantities with relatively lower tariffs.

With the trade war outbreak, both tariffs and export quantities went down, which made it very costly for the manufacturing Taishang to continue their production.

Meanwhile, when the scale of production by mainland Taishang decreases, the demand for the imported intermediate goods/components from Taiwan decreases as well.

This will lead to the downscaling of the Taiwan firms in Taiwan and reduce Taiwan’s exports to mainland China.

Due to the above reasons, we expect that the number of new TDI investment projects in the manufacturing sector in mainland China would decrease and the dollar amount of the TDI investment in mainland China would decrease as well, particularly with the outbreak of the third round of the trade war, given everything else held constant.

In comparison, since the Taishang in the nonmanufacturing sector are less tied to the ROC–PRC–US trade triangle, it is expected they would not be impacted much by the trade war.

Causal mechanism of the impact of Beijing’s Taishang-preferential Policies on Taishang during Trade War

During the trade war period, Taishang played an important signaling role for mainland China.

TDI mainland China seemed to serve as a positive signal for FDI from other source countries. The logic is that, given the tense and hostile relationship across the Strait, if the Taishang continue increasing their investment in their adversarial economy, it indicates a certain reliability of mainland China as an FDI-hosting countries.

To the same token, with the trade war escalating, if the Taishang had been withdrawing, despite their advanced and established industrial chain in mainland China, it would send a signal that mainland China is no longer a reliable or profitable location for FDI.

This can cause a domino effect and incentivize other FDI withdrawal from China as well.

Therefore, retaining the Taishang is not only important politically and strategically but also economically, considering the loss of TDI would snowball into further losses of FDI.

As a matter of fact, some serious Taishang incentive policies for retaining TDI in mainland China were issued before the outbreak of the US–China trade war.

Despite the tension across the Strait and the challenges Taishang are facing in mainland China in the post-2009 period, mainland China reemphasized its preferential policies toward Taishang at the beginning of 2018.

On 28 February 2018, the Mainland Taiwan Affairs Office under the State Council issued one of the most important Taishang-preferential policies, referred to as “measures to promote cross-strait economic and cultural exchanges and cooperation.”

This new policy has provided 31 Taishang-preferential policies to benefit Taishang in mainland China and has been termed as “the 31 Taishang-preferential Policies.”

These policies mainly affect Taishang and Taibao (i.e., the Taiwan people living in mainland China) in the fields of industry, fiscal policies, land-use policies, finance, employment, education, cultural exchanges, medical care, and so forth.

The major theme of the 31 Taishang-preferential policies was to grant the special status to Taishang and Taibao so that they could be treated equally with the mainland domestic firms and PRC citizens in mainland China.

The first 12 items on the 31 Taishang-preferential Policies are for the Taishang in mainland China, and the latter 19 items are for the Taibao.

The English translation of the Taishang-preferential policies (the first 12 items of the 31 Taishang-preferential Policies) are in the index.

The issuing of the 31 Taishang-preferential Policies aimed to reduce the cost of production for Taishang and assist with their industrial upgrading.

It also sought to help the Taishang to explore mainland markets and settle down in mainland China. Moreover, the policies also provide means to facilitate the Taibao to study, start business, find jobs, and settle down in mainland China.

The general goal of the 31 Taishang-preferential Policies is to increase the economic and cultural integration across the Strait.

It is important to investigate whether the 31 Taishang-preferential Policies played a positive role in attracting Taishang after the outbreak of the trade war.

The width and depth of this openness to Taishang and Taibao are unprecedented for mainland China, and Taishang and Taibao seem to find mainland China a very attractive location for conducting business and settling down.

Taiwanese society is concerned about the brain drain, as the nation’s elite, high-skilled human capital is flowing from Taiwan to mainland China to take advantage of the new policies under the 31 Taishang-preferential policies, particularly given the oversupply of high-skilled labor in Taiwan.

There are also concerns that the 31 Taishang-preferential Policies may change the identity of the Taishang and Taibao as they settle down in mainland China and enjoy such preferential treatment.

Following this logic, Taipei perceives the 31 Taishang-preferential Policies as another of Beijing’s mechanizations for cross-Strait reunification.

However, several months after the State Council’s issuing of the 31 Taishang-preferential Policies, the trade disputes broke out in early June and continued escalating until the end of 2018.

It is important to investigate whether the 31 Taishang-preferential Policies played a positive role in attracting Taishang after the outbreak of the trade war.

Although Beijing issued the 31 Taishang-preferential Policies in February 2018, it took the local governments some time to interpret and carry out these policies in their jurisdictions.

After all, all FDI are local, as such arrangements need to work with the local governments in China for their everyday business operations.

Most provinces in China have broken down the 31 Taishang-preferential Policies into more detailed measures and implemented their local versions of the policies at various times following their issuance.

For example, Shanghai started carrying out the policies on 29 May 2018, which was followed by Fujian province on 6 June 2018.

Other provinces started carrying out the policies later that year. Some provinces started before the trade war, and others after the trade war.

By the beginning of 2019, almost all the 31 provincial units in China had publicly issued their local version of the policies and announced the beginning of their application.

This provides us an ideal opportunity to investigate whether the issuing of the local versions of the 31 Taishang-preferential Policies played a positive role in attracting TDI inflows by holding all other social/economic factors constant for different provinces.

We expect that the 31 Taishang-preferential Policies would play a positive role in attracting Taishang in both the manufacturing and nonmanufacturing sectors.

Does the Outbreak of the Trade War Offset the Incentivizing Role of the 31 Taishang-preferential Policies?

According to the 31 Taishang-preferential Policies, the first (item 1, 2, and 3) and last three items (10, 11, and 12) highlighted the preferential policies granted to the Taiwan elites in the research-and-develpment field and the Taishang in the finance sector.

The Taishang in construction and agriculture sectors, as mentioned in items 4 and 9, are also provided with equal treatment status with the mainland counterparts.

Items 5, 6, 7, and 8 can apply to any Taishang in mainland China, including those in traditional manufacturing, which account for more than 75 percent of the total Taiwanese in mainland China.

These preferential policies toward Taishang in the manufacturing sector reduce the land-use fees for Taishang, prioritize Taishang in land-use approvals, allow them to make government purchases and form joint ventures with state-owned enterprises, and help the low-end manufacturing Taishang to relocate, upgrade, and take advantage of the Chinese market.

The temporal domain is from 1 January 2018 to 31 December 2018. The spatial domain includes 12 provinces whose Taishang data have been consistently recorded by the Investment Commission in the ROC.

In other words, although Taishang expected cost increases in producing in mainland China, some of the 31 Taishang-preferential Policies were designed to reduce the production cost for the Taishang.

Moreover, these policies are expected to increase the high-tech, service-oriented, and market-oriented Taishang to move to China.

Therefore, we anticipate that the 31 Taishang-preferential Policies may neutralize the negative impact of trade war on Taishang in mainland China.

Also, since the trade tensions across the Strait is still changing dynamically and it is very costly for Taishang to move out of mainland China given the unique and irreplaceable huge supply chain in China, we expect the 31 Taishang-preferential Policies may neutralize some of the negative impact of the trade war on Taishang.

Meanwhile, we expect that the strength of the marginal effect of Taishang-preferential policies has weakened with the outbreak of the trade war. Therefore, we made the third hypothesis in this article as follows.

Research Design

To investigate the above theory and hypotheses, we first conducted some descriptive analysis. Then we used the differences-in-differences (DID) with fixed effect estimation to empirically test the impact of trade war and the 31 Taishang-preferential Policies on Taishang in general—and manufactural and on nonmanufactural Taishang specifically.

Data Description

To test our theory and hypotheses, we utilized several datasets sourced from the ROC Ministry of Economic Affairs (MOEA) Investment Commission.

This source provides yearly and monthly TDI for Taishang at provincial level in China. The social and economic indicator data of the 31 provinces in China are sourced from the National Statistics Bureau of the PRC.

To empirically test the impact of the 31 Taishang-preferential Policies and the third wave of the trade war, we used the monthly data at the provincial level for 2018. Our unit of analysis is province–month.

The temporal domain is from 1 January 2018 to 31 December 2018. The spatial domain includes 12 provinces[14] whose Taishang data have been consistently recorded by the Investment Commission in the ROC.

Empirical Model and Measurements of Variables

To test our theory and hypotheses, we built the following empirical model:

Taishang_it = β₀ + β₁TradeWar3_it + β₂NewTaishangPolicies_it + β₃TradeWar3_it*NewTaishangPolicies_it + TradebyFDI_it + GDP at provincial level_it + Loss of profits by companies_it + Taiwan trade to mainland China_it + Construction areas for commerce_it + Retail commdity price index_it + Investment in fixed assets_it + Consumer Price Index_it + AddedValuesofIndustries_it + ᶓ_it

The first set of dependent variables includes the dollar amount (in 100 million USD) of (1) the monthly approved TDI in manufacturing sector; (2) the monthly approved TDI in the nonmanufacturing sector; and (3) the cumulative dollar amount of TDI in a given province up to a given month.

To correct the skewness of these dependent variables, we took natural log for them. Another set of dependent variables includes the number of cases/projects of (1) the monthly approved TDI manufacturing sector; (2) the monthly approved TDI in the nonmanufacturing sector; and (3) the cumulative number of Taishang cases in a given province up to a given month.

The data for all the dependent variables are from the ROC MOEA Investment Commission. The temporal domain of our data is from 1 January 2018 to 31 December 2018. The spatial domain of our data includes 31 provinces in China.

Table 1. Descriptive statistics of the variables

Variable	Obs	Mean	Std. Dev.	Min	Max
Cumulative amount of Taishang	116	13842.9	16088.2	412.4	55725.0
Cumulative cases of Taishang	116	3326.2	3906.7	102	13210
Monthly amount of Taishang	120	63.1	93.2	0	474.7
Monthly cases of Taishang	120	4.8	6.2	0	31
Third wave of trade war	116	0.4	0.5	0	1
New Taishang Policies	116	0.5	0.5	0	1
Monthly provincial GDP	116	10775.4	7598.5	2099.7	32425.9
Investment in industrial equipments	116	9.0	20.1	-26.9	99.4
Monthly export from Taiwan to China	116	8315.7	648.1	7285.5	9767.3
Profit losses by companies	116	12.9	24.9	-43.8	77.9
Construction areas for business	116	-4.7	3.6	-12.5	5.6
Retail commodity price index	116	101.8	0.7	99.8	103.9
Investment in fixed assets	116	6.8	6.1	-13.9	13.8
Consumer price index	116	102.0	0.5	101.1	103.4
Industry growth rate	116	5.6	4.3	-7	15.7
Month	116	7.6	2.9	3	12

The key independent variables include the trade war treatment, the 31 Taishang-preferential Policies treatment, and the interaction of these two treatments.

Trade war treatment variable is coded as “1” if the month is larger or equal to “9” and “0” otherwise. The reason for this coding is because the first two waves of trade war did not influence Taishang that much and the third wave of trade war impacted Taishang the most due to the large variety of goods on the tariff lists.

The 31 Taishang-preferential Policies treatment is coded as “1” if a given province started issuing and carrying out its local version of the 31 Taishang-preferential Policies, and “0” otherwise.

The interaction of these two treatments is created by multiplying them. We manually coded this variable based upon the official news reports presented by various provincial government websites.

The key control variables are all sourced from the National Statistics Bureau of the PRC, except that the Taiwan trade to mainland China data is sourced from the ROC MOEA.

The first control variable is a given province’s dependence upon export-oriented foreign investment, which is measured by the total trade made by the FDI in a given province in a given month.

We argue that the more a given province is dependent upon the export-oriented foreign investment, the more manufacturing Taishang it would attract, because most manufacturing Taishang are original equipment manufacturers (OEM).

The second control variable is the GDP at provincial level, which measures the economic development level in a given province in a given month. We argue that the more developed a given province is in a given month, the more Taishang it would attract.

The third control variable is the loss of profits by companies in a given province. We argue that the higher the loss of profits by companies in a given province, the less Taishang it would attract.

The fourth control variable is the dollar amount of trade from Taiwan to a given province in mainland China in a given year.

We argue that the more Taiwan trades with a given province in a given month, the more Taishang it would attract.

The fifth control variable is the construction areas for commerce, which measures how active the businesses are in a given province in a given month.

It is expected to have a positive impact on Taishang in a given province in a given month. The sixth control variable is the retail commodity price index, which is expected to have a positive impact on Taishang investment in mainland China, because the higher commodity price could lead to higher profits.

The seventh control variable is the consumer price index (CPI), which is expected to have positive impact on Taishang investment in mainland China, because the higher CPI, the higher level of living conditions in a given province in a given month.

The last control variable is the added value of industries in a given province in a given month, which is expected to have a positive impact on Taishang, because the high added value of industries indicates more profits and thus could attract more Taishang inflows. Table 1 shows the summary statistics for our variables.

Analysis

After conducting DID analysis, we made several findings based on the estimation results presented in tables 2 and 3.

Table 2 reports the results based on the dollar value of the TDI, and table 3 presents the results based on the number of Taishang projects.

According to table 2, under the situation where there is no issuing of the new Taishang policies, the third wave of the trade war did not have any statistically significant impact on cumulative TDI or monthly manufacturing TDI.

However, it does have negative and statistically significant impact on nonmanufacturing TDI. This finding is contrary to our hypotheses 1.1 and 1.2.

This indicates that the monthly changes of the newly approved TDI in the manufacturing sector is not significant.

This is possibly because the TDI in the manufacturing sector had already been decreasing over the past three years, and the outbreak of the trade war did not add much more negative impact on the downward trend.

However, the non-manufacturing TDI was increasing at the beginning of the 2018, and this upper trend was set back by the outbreak of the trade war.

Another important finding according to table 2 is that the new Taishang policies did not have significant impact on the monthly approved new Taishang inflows, but did have positive and statistically significant impact on the cumulative dollar amount of the Taishang in mainland China.

This indicates that although the new Taishang policies did not dramatically encourage the inflows of the newly approved Taishang, they did encourage added investment from existing Taishang in mainland China.

This indicates that the new Taishang policies strengthened the confidence of the existing Taishang in mainland China, but the policies’ impact on the new Taishang is still not apparent.

In addition, table 2 also indicates that the impact of the new policies did not change significantly with the outbreak of the third wave of the trade war for either manufacturing TDI or nonmanufacturing TDI.

This indicates that the resilience of the new Taishang policies during the post–trade war period.

Table 2. Differences-in-differences estimation results with dynamic linear panel regression on the dollar amount of Taishang in a given year in a given month

	Cumulative TDI	Manufacturing TDI	Nonmanufacturing TDI
Variable Names	Est. Coef.	Est. Coef.	Est. Coef.
Trade War 3	-0.1248 (0.1138)	5.575 (19.042)	-46.761* (26.331)
New Taishang policies	0.0067*** (0.0031)	-0.138 (0.964)	-1.057 (0.670)
Interaction	-0.0041 (0.0041)	1.061 (1.240)	1.341 (1.166)
Monthly trade by FDI	0.0006** (0.0003)	-0.272 (0.157*)	0.398** (0.179)
Monthly GDP	0.1427** (0.0723)	2.387*** (0.621)	1.472* (0.785)
Loss of profits by companies	-0.0002 (0.0001)	0.001 (0.013)	-0.003 (0.012)
Taiwan trade to mainland China	0.0484 (0.0631)	10.951 (22.039)	-50.773* (28.486)
Construction areas for commerce	0.0010 (0.0020)	0.143 (0.211)	-0.351 (0.270)
Retail commdity price index	-0.0050 (0.0038)	0.326 (0.440)	-0.297 (0.573)
Investment in fixed assets	0.0008 (0.0011)	0.056 (0.045)	-0.054* (0.033)
Consumer Price Index	0.0041 (0.0037)	0.281 (0.502)	-2.084*** (0.649)
Added values of industries	-0.0001 (0.0004)	-0.084 (0.068)	-0.022 (0.058)

Significance levels: *p<0.1 **p<0.05 ***p<0.01. Standard errors in this table are in the parenthesis and they are WC-Robust standard errors.

In addition to testing the dollar amount of TDI, we also tested the number of cases of Taishang in mainland China.

Table 3 shows the results for the cumulative cases of Taishang, the cases of manufacturing Taishang and the cases of the non-manufacturing Taishang, respectively.

The result shows that the outbreak of the trade war actually increased the number of newly approved manufacturing Taishang and increased the number of cumulative Taishang projects in mainland China, despite the trade war’s insignificant impact on the cases of the non-manufacturing Taishang.

This indicates that despite the risks faced by Taishang during the trade war, the attractiveness of mainland China as a potential investment destination remains.

Another possible reason is that the newly approved Taiwan are participating more in the “Red Supply Chain” to serve the domestic market within China.

However, the issuing of the new Taishang policies did not play a statistically significant role in attracting more Taishang projects as desired.

This is possibly because the time range under this study is still too short. Moreover, the impact of the new Taishang policies did not change much with the outbreak of the third wave of the trade war.

Table 3. Differences-in-differences estimation results with dynamic panel model for count data on the number of Taishang projects in a given year in a given month

	Cumulative Cases of Taishang	Cases of Manufacturing Taishang	Cases of Nonmanufacturing Taishang
	Est. Coef.	Est. Coef.	Est. Coef.
Trade War 3	0.0940* (0.0510)	9.172** (3.936)	-7.768 (7.957)
New Taishang policies	0.0005 (0.0007)	0.504 (0.465)	0.069 (0.178)
Interaction	-0.0004 (0.0015)	0.592 (1.026)	-0.058 (0.109)
Monthly trade by FDI	-0.0001 (0.0002)	-0.098 (0.117)	1.006** (0.413)
Monthly GDP	-0.0031 (0.0283)	1.291*** (0.385)	-0.002 (0.009)
Loss of profits by companies	-0.0001** (0.0000)	0.000 (0.007)	-6.638 (9.357)
Taiwan trade to mainland China	0.0871*** (0.0161)	13.555*** (4.369)	-0.050 (0.078)
Construction areas for commerce	0.0002 (0.0004)	0.082 (0.086)	-0.007 (0.274)
Retail commodity price index	0.0016 (0.0014)	0.229 (0.273)	0.005 (0.020)
Investment in fixed assets	0.0007 (0.0006)	-0.011 (0.034)	0.010 (0.590)
Consumer price index	0.0024 (0.0019)	-0.056 (0.322)	-0.044 (0.039)
Added values of industries	-0.0003** (0.0001)	0.003 (0.040)	-0.044 (0.039)

Significance levels: *p<0.1 **p<0.05 ***p<0.01. Standard errors in this table are in the parenthesis and they are WC-Robust standard errors.

Summary and Conclusion

Despite the economic risks under the third wave of the trade war and the political risks across the Taiwan Strait, many Taishang decided to stay, and some new Taishang continued to move to mainland China, which makes the total number of Taishang in mainland China increase in the recent several years.

Our empirical findings show that the third wave of the trade war did not have any statistically significant impact on cumulative TDI or monthly manufacturing TDI, despite some negative impact on the nonmanufacturing TDI.

Moreover, our empirical findings show that the outbreak of the trade war actually had a positive impact on the cumulative existing Taishang projects in mainland and on the number of monthly newly approved Taishang projects in mainland China.

In addition, our study also found that Beijing’s new Taishang policies have already started playing a positive role in encouraging more reinvestment by the existing Taishang in mainland China, although it has not shown any significant impact on the monthly approved new Taishang in mainland China.

Meanwhile, the marginal effect of the new policies did not strengthen or weaken with the outbreak of the trade war.

This indicates the possible resilience of the new policies. Another possible explanation is that it will take a little longer for the new policies to show a positive effect.

It is important to investigate the Taishang issue, since these businesspeople are important assets for the PRC in both the economic and political sense.

The exit of Taishang would not only weaken the communication channel across the Strait but also reduce the PRC’s leverage upon Taiwan.

Thus, Beijing issued the 31 Taishang-preferential Policies, which aim to incentivize Taishang to continue investing in mainland China.

Following the central government’s policies, the local governments in China also carried out their own versions of the policies, aiming to retain the Taishang in their jurisdictions and attract new Taishang to flow in.

For future studies, it is worth continuous investigation as to how effective these policies are. In addition, more work needs to be done to expand the temporal domain of this study, as the trade war continues developing and more data will become available.

The major weakness of focusing on the 12 months of the year 2018 is that the impact of the trade war may not have taken effect given the short time range.

We will continue collecting the monthly data for Taishang and conduct further analysis with the development of the trade war.

Finally, we will incorporate the ROC’s Taishang incentive policies in our future statistically models as well.

(Dr. Kelan “Lilly” Lu is an assistant professor in the Department of Political Science at the University of South Carolina; and Mr. Kuan-Wu Chen is a graduate student in the Department of Political Science at the University of South Carolina)