2.1 Trademark Claims of SMEs and the Underlying Grounds

Scholars have long studied the rationale underlying trademark claims brought by plaintiffs, which has facilitated the development of common law and statutes. One important concern raised is the distinction between ‘harm’ (caused to plaintiffs) and ‘legitimate interests’ (enjoyed by plaintiffs). Plaintiffs tend to ignore a crucial point, especially in post-sale confusion cases: plaintiffs should prove the injury occurred to the consumer rather than only claim that somebody else profited from the trading (Lemley and McKenna, Reference Lemley and McKenna2010, p. 188). Trademark law aims to prevent consumer confusion and to ask whether consumers are confused by the defendant's use of the mark (Lemley, Reference Lemley2004). The problems with post-sale confusion cases are that consumers could acquire the prestige value by buying fakes or cheap imitations (Lemley and McKenna, Reference Lemley and McKenna2010, p. 152). The plaintiff in a post-sale confusion case thus must show that either it suffers injury from the source confusion or the defendant's use arouses other types of confusion that are material to consumer purchasing decisions (pp. 156–188). Meeting this burden would justify the plaintiff's claim of harm. Otherwise, this situation would be a ‘trademark injury,’ affecting the trademark owner's trademark rights instead of causing harm (p. 188).

Other scholars have paid attention to the confusion imposed on consumers and the infringing activities that affect them in acting on their preferences rather than looking at how defendants misuse the meaning of the plaintiff's mark (McKenna, Reference McKenna2012, p. 132). Tushnet (Reference Tushnet2016, p. 629) argued that in many circumstances if there is no harm to consumers, there should be no harm to trademark owners. Even though the case may involve harm to consumers, such harm is always separate from the injury to trademark owners (Tushnet, Reference Tushnet2016, p. 636). Multinational companies claim significant losses due to counterfeits in developing countries, but such claimsrepresent only a loss of control in how they present their brand, which is not a financial issue (Chow, Reference Chow2010, p. 822). The exaggerated financial loss claims of multinational companies are not justified by scientific evidence or the calculation of damages. These companies are only trying to eliminate the enforcement cost/responsibilities in developing countries that host manufacturing (p. 823).

2.2 Direct Competition, Dilution Claims, and Unfair Competition

Based on the rationale of trademark infringement, the very fundamental question in granting compensation to plaintiffs should be whether there is direct competition. Without direct competition, there should be no economic justification to compensate plaintiffs for lost profits on diverted sales (Corgill, Reference Corgill1996, p. 1922). Specifically, the reputation of a trademark represents the investments and costs incurred to trademark owners. In this sense, trademark infringers avoid these fixed costs and benefit from the sales to consumers (1938–1940). Therefore, before granting remedies to a plaintiff, we need to understand what costs infringers avoid and whether the plaintiff bore such costs. Additionally, Buccafusco and Masur (Reference Buccafusco and Masur2013) focused on the balance between free-riding and free competition. Looking at the ‘form’ and the ‘nature’ of the infringing activities, they asserted that the nature of the infringement should involve ‘direct substitutes’ and ‘exact duplication’ works or products (p. 316).

With the development of trademark theories, scholars have argued that current sanctions appear too harsh for infringements that do not harm consumers or trademark owners. For example, when a consumer knowingly purchases a counterfeit Rolex, there has been no harm to the consumer because there is no injury (Rierson, Reference Rierson2007, p. 435). In other words, whether the trademark owner suffers a loss of sale depends on whether the consumer has been deceived (Bartow, Reference Bartow2011, p. 742). We must pay attention to the nature and degree of deception of consumers in counterfeiting cases. Meanwhile, when defining ‘infringement’, we need to understand whether the defendant's action would affect the plaintiff's market (Fromer and Lemley, Reference Fromer and Lemley2013). However, this theory is criticized because, in many cases, consumers probably pay more attention to the function of the products rather than the brand attached to them (Lee et al., Reference McKenna2009, p. 1099). In this sense, the ‘market criteria’ are easily met, while consumers are not deceived at all. This would lead to an ‘easy infringement’, further reducing the options on the market (Fromer, 2013, p. 1277; Tom et al., Reference Tom, Garibaldi, Zeng and Pilcher1998).

Given their increased communitive function, trademarks are more often used to transmit brand value, images, and consumer associations. Thus, trademark loss may not be limited to direct competition. For example, fake watches may not be direct competitors to Rolex, but Rolex may still be unhappy about counterfeits. The harm to Rolex could be dilution or erosion of its brand reputation.

In 1989, the European Trade Marks Directive (TMD) introduced and allowed for the protection of trademarks with a reputation against harm to distinctives and unfair advantages.Footnote ⁷ With the theory of dilutive harm, non-confusing uses of a famous registered mark can dilute their selling power, eventually leading to a loss of distinctive character and the demise of the mark (Schechter, Reference Schechter1927, p. 825). Especially within the luxury industry, consumers receive both a tangible product and an intangible added value (Assaf, Reference Assaf2009, p. 643). Anti-dilution protection extends the trademark protection (from anti-confusing uses) to brand communication, advertisement, and investment (Welkowitz, Reference Welkowitz2012; Kur and Senftleben, Reference Kur and Senftleben2017, p. 183).

In contrast to traditional trademark law, which looks to actual economic harm, anti-dilution protection pays attention to the trademark owners’ loss of investment and the impairment of the brand's selling power (Long, Reference Long2006, p. 1034). Though the explanations of dilutive harm have long been accepted in both the US and Europe, scholars have criticized the vaguely constructed wording in regulatory articles and case law. They argued that preserving the uniqueness of a mark indeed ignores the misappropriation rationale underlying the theory (Bone, Reference Bone2007, p. 471; Rierson, Reference Rierson2012, p. 246).

Besides dilution, unfair competition serves as another rationale for trademark protection. At the core of unfair competition is that anyone who uses another's trademark should not be legally permitted to unfairly profit from this use. Supported by unjust enrichment (Barnes, Reference Barnes2010, p. 470) and Lockean labor theory, this rationale aims to prevent the misappropriation or free-riding of someone else's reputation, effort, or time without their consent (Griffiths, Reference Griffiths2007, p. 138). Though such a catch-all approach is criticized for being unconfined for many scholars (Dogan, Reference Dogan2006, p. 103), it represents the extension of trademark theories to incorporate both consumer deception and all the interests worth protection.

2.3 Factors Considered in Judicial Decisions and Remedy Calculations

In civil litigations, only a few factors drive the case outcome, so paying more attention to them seems to be economical (Stigler, Reference Stigler1970; Gigerenzer and Goldstein, Reference Gigerenzer and Goldstein1996; Simon, Reference Simon2004, p. 1603). The decision on damages relies on the facts of the case. In trademark cases, the facts are the type of product involved, the litigating parties, and the power of the mark. This is especially true in counterfeiting cases because almost all product types have counterfeits in a wide range of industries (Wilson et al., Reference Wilson, Brandon and Hollis2016, pp. 259–260).

These perspectives can integrate into a scheme that reflects all available information about counterfeiting. Courts traditionally rely on their understanding of the facts and statutes to decide how much monetary relief to grant. Given the complexity and technological content of intellectual property (IP) cases, many regimes have included experts to answer questions related to fact-finding (e.g., consumer perspectives in trademark cases; Abramson, Reference Abramson2016, pp. 1097–1101).

Regarding damages and other reliefs granted by courts, Shapiro (Reference Shapiro2011) pointed out that punitive damages are ineffective in intentional wrongdoings. Punitive damages have fewer advantages than multiple damages and the drawback of calculation difficulties (Shapiro, 2010, p. 498). Shapiro believes that multiple damages may work better than punitive damages because multiple damages can target those intentional activities, such as trademark counterfeiting, that are truly intentional and worth deterrence but are not reprehensible; while punitive damages target those crimes considered more reprehensible (pp. 460–488).

2.4 Problems in China's Legal Mechanism Regarding Civil Remedies

Over recent decades, the number of enforcement cases in China has increased under external pressures from international society. However, such an increase in cases does not necessarily represent an increase in the quality of adjudications or enforcement (Dimitrov, Reference Dimitrov2009, pp. 9–26). Dimitrov (Reference Dimitrov2009) generated a complete picture of China's overall mechanism of dealing with IP infringement. Though he did not provide detailed empirical analysis or causal inference on why enforcement and adjudications appear to be low and ineffective, Dimitrov showed the systematic problems in local enforcement and the Chinese judiciary. He suggested enforcement improvements, including targeting custom border control and encouraging local administrative detection. Similarly, Brauer (Reference Brauer2012, pp. 277–290) asserted that even though laws are settled on paper, in reality, civil litigation and anti-dilution protection remain weak and unreliable for trademark owners.

Sepetys and Cox (Reference Sepetys and Cox2009) claimed that external pressures on China's IP legal regime primarily come from international treaties and complaints from other countries with many IP products entering the Chinese market. IP legislation usually causes a loss to consumers in the form of higher prices, but this cost is counterbalanced by social innovation and invention. From the perspective of civil remedies in judicial decisions, Sepetys and Cox also found that IP infringement cases in China have a low-unit price in each case and lack records of transactions (by defendants). These two problems result in a low compensation amount and difficulties in determining the illegal gains of infringers (pp. 6–7). The two most startling findings in Sepetys and Cox concerned monetary reliefs in Chinese IP cases: (1) the median award is about USD 15,000, and 90% of cases receive less than USD 100,000, and (2) Chinese plaintiffs receive less than foreign plaintiffs.

In studies of trademark infringement in China, judicial decisions have received less attention than statutory/doctrinal studies because of the country's civil law system. Through their case analysis, Sepetys and Cox (Reference Sepetys and Cox2009) provided detailed examination and calculation before drawing a general conclusion on civil IP protection in China. Though they did not study each type of IP, their work raised the possibility of exploring a civil law regime.

2.5 Empirical Findings in Chinese IP Studies

Scholars in China have explored, through empirical analysis, the civil protections and remedies in IP cases. Their recommendations for legislative reform fall into a few aspects. First, and consistent with scholars elsewhere, they found that the mean remedies granted in trademark infringement cases total around USD 12,000, with almost no changes between 2005 and 2011 (Duan and Xie, Reference Duan and Xie2014, p. 15). This finding was a remarkable progress in the empirical studies of judicial trademark decisions in China. But the conclusions on the decision trends (which remain unchanged) need further discussions: the remedies look unchanged may be due to the non-categorizing nature of the studies. These studies do not categorize the cause of action (e.g., trademark counterfeiting or general infringement, trademark or geographical indication not given) and only look at the overall situation of cases.

Second, and related to judicial calculation of damages, some scholars have pointed out that statutory damages are used too frequently (Duan and Xie, Reference Duan and Xie2014, pp. 21–23). The overuse of statutory damages causes low-quality decisions. Those cases involving more technical elements and foreign litigators are more likely to receive detailed and reasoned decisions (Lan, Reference Lan2012).

Third, some scholars have proposed a new mechanism to deal with cases involving civil litigation and criminal violations (Zhang, Reference Zhang2006; Sun and Dong, Reference Sun and Dong2007). In cases involving civil litigation and criminal violations, the criminal case traditionally takes priority. However, this traditional mechanism may not work for IP cases, and a mechanism in which civil litigation enjoys priority should be considered (Sun and Dong, Reference Sun and Dong2007, p. 591). Unlike torts and other crimes, IP crimes (especially trademark counterfeiting) are defined by the infringement amount. However, if the case fails to reach a monetary threshold, defendants may avoid criminal prosecution and only face a civil suit. Thus, prioritizing civil cases would assist in determining whether infringement exists and the seriousness of the liability or crime. Furthermore, IP rights are, in nature, private rights. Once defendants are found liable and pay the monetary remedies, they do not pose harm to the public or society, which may save judicial resources in avoiding criminal prosecution (596–597).

This section has reviewed the literatu on Chinese judicial practice in IP cases – though not specifically trademark – and spotted problems in the procedures. The review indicates positive trends: the legislation on IP infringement is working toward a better mechanism, and the judiciary has become aware of the drawbacks in the overall regime design.

3.1 Reason for Empirically Studying Counterfeiting Issues in China

Previous studies discuss trademark statutes and the justifications of civil claims. Most studies focus broadly on trademark infringement without distinguishing trademark counterfeiting cases from infringement cases or theoretically look at the civil protection by courts and agencies. Few studies have looked empirically at civil litigation in trademark counterfeiting – even though it is the most serious type of infringement, and there could be criminal liability if the threshold is reached. Meanwhile, IP scholars have asserted that counterfeiting comprises many factors, such as trademark owners/entities, counterfeiters, and products counterfeited. Discussions of these factors within the context of a ‘bigger picture’ have been fragmented throughout the years. A broader scheme that reflects as many factors as possible is needed to show the different aspects of trademark counterfeiting.

The case study is not an innovation in legal research. Plenty of IP scholarship has employed the case study and further generated empirical analysis or policy recommendations. But in trademark counterfeiting, case studies on civil judicial decisions are rare, probably because counterfeiting is often a low priority for prosecution agencies (Wilson et al., Reference Wilson, Brandon and Hollis2016, p. 270). From an international trade and policy perspective, we are interested in whether counterfeiting enforcement in China has improved with the long-lasting external pressure. We also want to know the suggestions regarding the litigations of SMEs. We try to explore the characteristics and trends of trademark counterfeiting litigations from China's perspective, employing a detailed empirical study of government annual reports and case studies.

3.1.1 Data

This study relies on the 2011–2015 annual reports of the Supreme Court of the PRC,Footnote ⁸ which discloses the number of trademark case-closed rates for each year. Among all trademark civil decisions, counterfeiting cases comprise 27.9% of all trademark cases (26525), with a total of 7,388 (population). The data came from the most widely used legal research database in China: PKULaw.Footnote ⁹ For the overall dataset, 826 cases were coded, comprising 10.7% of the population.Footnote ¹⁰

The data selection and analysis consisted of two stages. In the first stage, 401 cases were randomly selected to conduct a descriptive analysis and generate a ‘big picture’ of civil decisions on trademark counterfeiting in China. In the second stage, 425 cases were added to confirm findings in the first stage and further develop a coding system with which to analyze civil litigation. Table 1 presents the closed trademark civil cases in each year and the dataset coded.

Table 1. Number of cases by year (all data)

Notably, under Chinese civil procedure, the retroactive application of the 2013 law revision depends on when the counterfeiting activities ended and when the civil litigation began. The new law does not retroactively apply to civil cases litigated before it took effect. Similarly, if the counterfeiting activities ended before the new law took effect and the plaintiffs litigated after it entered into force, the new law still did not apply. However, if the counterfeiting activities lasted until 2013, when the new law took effect, then the new law applied. Therefore, this study coded cases according to the year of closing and when the counterfeiting activities occurred. The year coded for each case here indicates that the law had been applied (for further analysis).

3.2 Regression Model Designs, Assumptions, and Coding System

Through ordinary least squares (OLS) regressions, this study tries to explore the association between remedies (outcome variable) and the various characteristics of parties and cases (independent variables). This helps to illustrate what factors drive remedies higher than others. The assumption and model established here is that the amount of monetary remedies is associated with how much plaintiffs claimed, whether the case had a prior decision, if the court provided a calculation of remedies or legal reasoning, whether the court granted an injunction, what type of company the plaintiff and defendant were, the type of product, and if the case was decided after the law revision.

Furthermore, this study also aims to find out the relationship between monetary relief and the grant of an injunction after the 2013 law revision. After the law change, did the monetary relief become greater for every grant of injunction than before the law changed? The study sought to compare the monetary relief in three periods: 3 months after the law was revised; 7 months after the law was revised; and all cases in the 28 month period after the law was revised (September 2013–December 2015) – and then compare them with the monetary remedies granted before the law revision. The assumption established for the time effects and coding table with an interaction term are as below:

Most importantly, many previous studies on general IP judicial decisions do not differentiate among trademarks, patents, and copyright (Park and Ginarte, Reference Park and Ginarte1997; Sepetys and Cox, Reference Sepetys and Cox2009). Without studying the law applications, these studies only look at the information shown in the decisions. To find out what causes some remedies to be higher than others, this study substantially researched into the judicial reasonings. Below, an elaboration of the factors and reasons for establishing these assumptions are explained (Table 2).

Table 2. Variables for OLS interaction effect of injunction and time on monetary relief

First, this study reviewed every case in detail to see whether the courts provided any reasons for their decisions and, if so, what those were. Scholars have pointed out that a large percentage of the cases in Chinese courts lacked reasoning (Cheung, Reference Cheung2009; Dimitrov, Reference Dimitrov2009). The courts only copied the legal clause in determining the remedy amount without illustration. In rare cases, the courts provided a detailed calculation based on the number of items and retail price. Therefore, the cases were coded into categories according to the extent of the judicial reasoning to test the assumption of whether remedies are associated with courts’ reasoning, and if they were, how.

Next, the powers of litigating parties sometimes impacted judicial decisions because large companies may have better legal resources or be better at document preparation. Thus, the plaintiffs were coded by type, such as foreign companies, state-owned companies, and domestic SMEs (referred to as ‘grassroot’ companies).Footnote ¹¹ Meanwhile, previous studies have found that most counterfeiters are family-style small factories or individuals (Cai, Reference Cai2020). This study paid attention to secondary liability legislation and coded whether the defendant company was a flea market, supermarket, or shopping mall to test the assumption of how party types relate to the remedies granted.

To compare remedies before and after the 2013 law revision, the year the case closed was coded in each case. The new trademark law took effect in August 2013. Thus, the time period 2011–August 2013 was treated as the time before the change, and the cases thereafter were treated as those after the change. Finally, Table 3 summarizes the variables coded and prepared for analysis.

Table 3. Explanation of Coding

4.1 How Much Courts Grant and Who Are the Litigating Parties?

The 2013 law revision increased the statutory damage cap from RMB 500,000 (USD 71,000) to RMB 3,000,000 (USD 428,000). Courts use their discretion to determine whether the plaintiff should prevail, such as the power of the evidence, the arguments of parties, and the justifiability of the defenses. Therefore, after the increase in the statutory damage cap, do courts grant more monetary remedies to plaintiffs? Are there any trends in the remedies for trademark counterfeiting cases? Table 4 provides the monetary remedies by summary statistics.

Table 4. Monetary Remedies by Year (in RMB ¥)

Monetary remedies, in most cases, remained reasonably stable during the 2011–2015 period. The annual mean remained stable at around RMB 42,000 (USD 6,000), with a slight upward trend after 2014. Higher remedies (i.e., max remedy) showed a slight increase in 2014 and 2015. The annual median reached its lowest point in 2013, rebounded in 2014, and then settled with a slight downward trajectory in 2015. Remedies for most of the cases are less than or equal to RMB 56,000 (USD 8,000), looking from the 75% distribution.

Besides examining the overall monetary remedies granted in trademark counterfeiting cases each year, this study also looked at the month-by-month civil remedies to generate a ‘bigger picture’ of the overall protection of trademark rights. It paid special attention to when the 2013 law revision took effect to see whether any changes occurred in civil monetary relief. The results are illustrated in Figure 1.

Figure 1. Monetary relief and injunction, by month

After the 2013 law revision took effect, monthly medians immediately increased for a few months. This change suggests that, first, the overall monetary remedies increased, and, second, more cases appear to have received significantly large monetary remedies. Additionally, the injunction rate in each month remained relatively stable in the 5-year time frame (around or above 80%). After the law revision took effect, the injunction rate did not change dramatically, either right after the revision or in later months. The date the law revision took effect, 30 August 2013, is clearly marked.

This study further looks at the remedies granted to five types of plaintiffs: foreign companies, state-owned companies, domestic public companies or famous private companies (domestic giants), joint ventures (JVs), and general SMEs (grassroot). Table 5 shows the monetary remedies granted to different types of plaintiffs. Large remedies (median and max) were granted to foreign and grassroot companies the most. The medians and means of foreign and grassroot companies are higher than others. The JVs appear to have the lowest mean and median.

Table 5. Monetary remedies granted by plaintiff types (in RMB ¥)

In comparison, do the damages paid by defendants differ across types? Do company defendants pay more damages than other defendants? Do flea market defendants pay more damages than individuals or retail stores? Flea markets may not have a commercial scale as large as a company, but their financial capability should be no less than that of an individual or retail store. Table 6 illustrates the remedies paid by defendant type.

Table 6. Monetary remedies by defendant types (in RMB ¥)

Table 6 shows that company defendants pay large remedies, followed by individuals and shopping malls. Trivial remedies appear most in flea markets (lowest mean and median), followed by individuals and shopping malls. The medians among company defendants are higher than those of shopping malls and flea markets, with company defendants having the highest mean.

4.2 Cases with Larger Remedies

The previous subsection explored the nature of litigating parties to study the differences in remedies across company type. This section explores the monetary remedies and distributions in cases where the courts explain their reasonings in granting a specific remedy amount.

Most courts do not provide explanations of how they decided the remedy amount (82%) as shown in Table 7. When we try to compare the remedies granted in the cases with no reasonings (82%) to those with some reasonings (15%) in Table 7, we notice that both the mean and median of cases with some reasonings (10715, 4286) are twice of the mean and median of those with no reasonings (5315, 1857). For those cases with detailed reasoning on the remedies granted (3%), the mean and median are not significantly larger than those with some explanations. However, the distribution appears to be trending upward, which indicates more large remedies. The 90th percentile cases with detailed reasonings are 1.6 times more than those with some explanation and 3.5 times more than those without any explanation.

Table 7. Decisions with court reasonings

4.3 Factors Associated with Greater Monetary Relief

This study explores in detail the factors associated with higher monetary remedies. It employs a linear regression model to predict the judicial monetary remedies as a function of the characteristics of the cases and litigating parties. The sample studied here consists of 788 cases with monetary remedies. The model posits that the higher monetary remedies (in RMB ¥) may relate to the detail of the judicial reasoning (calculation), whether an injunction is granted, what type of entities are the plaintiff (ptype) and defendant (dtype), and the year the case at issue was decide.

As Table 8 reflects, judicial monetary remedies vary in cases with prior decisions, judicial explanations, different types of litigating parties, and product types. Specifically, cases with a brief judicial explanation of factors considered (calculation-1) are granted RMB 16,000 (USD 2,300) more than cases without any reasoning. Cases with detailed judicial reasoning (calculation-2) lead to RMB 21,000 (USD 3,000) more in monetary relief than those without. In cases where monetary remedies are granted together with an injunction, the monetary remedy shows RMB 14,000 (USD 2,000) higher than cases without an injunction. Foreign plaintiffs (ptype-1) receive RMB 23,000 (USD 3,200) more than JVs, and company defendants (dtype-1) are more likely to pay higher damages (RMB 33,000 or USD 4,700) than individuals. Other types of plaintiffs (i.e., state-owned, grassroot, domestic giants) do not receive significantly higher remedies than JVs. Other types of defendants (i.e., flea markets, shopping malls) do not pay significantly more damages than individuals.

Table 8. Estimated Association between Factors and Monetary Remedies (in ¥)

Note: Ordinary least squares linear regression model on monetary remedies. ‘calculation’ = whether courts provide (a) considerations on some factors but no details or (b) detailed explanation on remedy calculation. Coefficient provided. *p < 0.1; **p < .05; ***p < .01.

Specifically, cases with a brief judicial explanation of factors considered (calculation-1) are granted RMB 16,000 (USD 2,300) more than cases without any reasonings. Cases with detailed judicial reasonings (calculation-2) lead to RMB 21,000 (USD 3,000) more in monetary relief than those without. A grant of injunction is associated with RMB 14,000 (USD 2,000) more in remedies than no injunction. Foreign plaintiffs (ptype-1) receive RMB 23,000 (USD 3,200) more than JVs, and company defendants (dtype-1) are more likely to pay higher damages (RMB 33,000 or USD 4,700) than individuals. Other types of plaintiffs (i.e., state-owned, grassroot, domestic giants) do not receive significantly higher remedies than JVs. Other types of defendants (i.e., flea markets, shopping malls) do not pay significantly more damages than individuals.

4.4 The Monetary Relief Changed After the Law Revision

Now, we look in more detail at whether the monetary relief changed after the law revision took effect in August 2013. As shown in Table 9, this study codes the month in which the case was decided. Table 9 shows that a few months immediately after the law revision took effect, there was an immediate upward trend in both of the monthly median of civil remedies. To see if the increase is significant compared with the months before the revision, Models A2 and A3 consider the three months after the revision to the first highest point of monthly mean – September–November 2013 – as ‘after revision’ cases (3-month span). Models A4 and A5 consider September 2013–March 2014 as after revision cases (7-month span). Finally, all cases from September 2013 to the end of the data period (1.5-year span) are considered in Models A6 and A7.

Table 9. Estimated Association between Interaction on Monetary Remedies

Note: Ordinary least squares linear regression model on monetary remedies. ‘September–November’; ‘September–March’; ‘September–end = cases closed during 2013.09–11; 2013.09–2014.03; 2013.09–2015.12, are treated as after revision respectively.

*p < 0.1; **p < .05; ***p < .01.

As Table 9 reflects, cases decided after the law revision in August 2013 are associated with higher monetary relief than cases before it. Cases decided September–November 2013 received RMB 37,000 (USD 5,200) more than cases prior (model_2–3). Expanding the time frame to March 2014, cases received RMB 22,000 (USD 3,000) higher than cases before the revision (model_4–5). All cases closed from September to the end of 2015, overall cases after the revision received RMB 30,000 (USD 4,000) higher than the previous cases (model_6–7). Furthermore, the months immediately after the law revision took effect – September 2013–November 2013 – have stronger positive effects than other periods.