Unintended consequences of National centralized volume-based drug procurement policy in China: a game theory-based analytical approach for pharmaceutical quality and pricing

Abstract

Background

The pharmaceutical industry plays a pivotal role in the healthcare economy, with drug quality and pricing being critical to consumer welfare. China’s government has implemented National centralized volume-based drug procurement(NVBP) policy, aiming to reduce medication costs through bulk purchasing. While existing research confirms its price-reduction effects, unintended consequences regarding pharmaceutical quality incentives remain overlooked. This study investigates how the NVBP policy reshapes firms’ quality-pricing strategies, addressing a critical gap in policy evaluation.

Methods

Game-theoretical model is constructed to simulate interactions between two pharmaceutical firms with distinct technological capabilities before and after China’s National volume-based procurement (NVBP) policy. Prior to the policy, the Hotelling model’s spatial competition is reinterpreted as a quality-preference spectrum After the policy, drugs are standardized through consistency evaluation, and competition reduces to pure price competition as in the Bertrand model. Then a double-envelope bidding game is introduced where the government sets a minimum quality standard and firms bid on price. After winning the bid, firms need to decide the production strategy.

Results

Before NVBP policy implementation, equal-quality price differences depend solely on technical parameters. Under endogenous quality, a firm’s quality rises with its own technology; the technologically superior firm charges a higher price, but a weaker firm facing a superior rival lowers its price. After NVBP policy implementation, during the bidding time, if both firms pass consistency evaluation, the winning bid price falls, but quality may drop below pre-NVBP levels. If the low-tech firm fails, the high-tech firm gains monopoly pricing power and raises prices. During the production time, some unintended consequences may occur, such as quality risk and supply disruption risk.

Conclusions

The implementation of NVBP policy could lower drug prices and improve access, but it may also generate unintended consequences. Low standards induce price competition and reduce quality investment, while high standards may create sole-supplier monopoly. Excessively low bids pose quality or supply risks. Policy should calibrate standards, use tiered incentives, and strengthen surveillance, but empirical validation remains needed.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12962-026-00794-1.

Background

The pharmaceutical industry plays a pivotal role in healthcare economies, with drug quality and pricing serving as critical determinants of consumer welfare and systemic sustainability. Historically, high drug prices have exacerbated the social challenge of “inaccessible healthcare due to cost barriers” [1, 2]. In response, China’s healthcare reform introduced price containment measures in 2019, and the State Council launched nationwide centralized drug procurement pilot policy to address price inflation and reduce public medication costs through market-oriented procurement mechanisms [3, 4]. Conventional drug pricing operates within competitive market frameworks where equilibrium emerges from supply-demand interactions [5, 6], with pharmaceutical firms strategically investing in R&D to enhance drug quality as a core competitive advantage [7]. As researchers noted, superior quality not only addresses unmet medical needs but also fosters patient trust and corporate reputation [8]. Meanwhile, pricing integrates costs from R&D, manufacturing, and distribution alongside technological innovation that reshapes therapeutic landscapes [9]. Regulatory frameworks for approvals, patents, and quality standards further modulate market behavior [10, 11], with firms prioritizing improvements in quality and marketing cost to maintain leadership—attributes directly linked to efficacy and safety [12].

China’s National Volume-based Procurement (NVBP) policy has fundamentally transformed market dynamics by replacing decentralized pricing with centralized tender to leverage consolidated buyer power for cost containment [13, 14]. Empirical evidence confirmed its effectiveness in reducing prices by 54% on average [15]; however, this shift has forced firms to reallocate R&D resources and optimize product portfolios. The policy’s “double-envelope” bidding system, which combines corporate credit, price and technical evaluations, creates unique strategic incentives. While recent studies have analyzed supply chain adjustments under this system but overlooked quality-strategy interactions and highlighted potential trade-offs between cost control and quality maintenance [16]. Existing research predominantly concentrates on three areas: price reduction effects [5, 17, 18], reshaping China’s pharmaceutical industry [19], and some problems, like patient trust and drug waste, etc [20]. About price reduction effects, research focuses on the variations in price reductions across different drug categories, specifically comparing formulation types and the gap between generics and originator brands [21]. About pharmaceutical industry, empirical evidence indicates that policy could foster R&D investment and innovation, with effects being particularly pronounced among firms characterized by high initial marketing expenditure and successful bidding performance. Furthermore, the policy has increased market concentration and optimized the industrial structure [22]. About problems caused by policy, the phenomenon of surging consumption and expenditure on alternative drugs (particularly original drugs in city-level hospitals) is attributed to manufacturers’ cost-shifting behaviors under squeezed profit margins, compounded by the inadequate clinical efficacy of bid-winning drugs, poor patient adherence, and physicians’ profit-driven prescribing of costlier equivalents [22].

While current studies have documented the impacts of NVBP policies on stakeholders such as pharmaceutical firms, the industry, and patients, few studies have yet examined these effects through the lens of overall market game theory. A critical gap remains: how do firms with heterogeneous technological capabilities adjust quality and pricing strategies under volume-based procurement? Prior studies either use homogeneous firm models [23] or lack game-theoretic analysis of quality-price trade-offs [24, 25]. They all fail to address the “quality dilution” and “monopoly risk” paradoxes under different quality standards. This study bridges this gap by constructing a game model incorporating China’s procurement mechanisms and firms’ technological differences, analyzing how low/high quality standards influence market equilibrium, quality outcomes, and pricing strategies, and proposing policy recommendations to balance cost control with quality assurance. By integrating technological heterogeneity and strategic interactions, the research provides theoretical insights for optimizing procurement policies in transitional healthcare markets, addressing both short-term cost containment and long-term innovation incentives.

Methods

We employed a differentiated duopoly game model with a contrast static analysis incorporating policy shocks to examine firms’ price competition before the implementation of the NVBP policy and their quality-price trade-offs after implementation. It comprises three stages: first, describing the problem and establishing fundamental assumptions; second, constructing equilibrium states in a two-stage game model before the policy implementation; and third, developing a bidding game model for firms after the policy implementation.

Model description

The volume-based drug procurement policy has been proven effective in reducing drug prices and alleviating patient burden through economies of scale in bulk purchasing in many countries. However, it remains unclear how governments can safeguard against pharmaceutical quality degradation or the emergence of new price collusion when using administrative measures to lower drug prices. Game theory is widely regarded as an essential tool for functioning analysis [26, 27]. There are many models of products differentiation, such as Hotelling model, Bertrand model, Salop circle model, etc. The Hotelling model is a classic model in industrial organization theory used to analyze spatial competition and product differentiation among firms. It aims to explain how firms compete based on location (physical or abstract) and price. The model typically employs a two-stage dynamic game to explain this: in the first stage, firms compete on location, and in the second stage, they compete on price [28]. The Bertrand model posits two firms offering identical products engaging in simultaneous price competition. It proves that equilibrium prices collapse to marginal cost, resulting in zero profits and underscoring the necessity of product differentiation or capacity limits for maintaining market power [29].

This study aims to characterize the quality and price of pharmaceutical products from the aspect of pharmaceutical manufacturers before and after the policy implementation. Prior to the implementation of NVBP policy, the pharmaceutical market was characterized by brand and quality differentiation. Firms competed for market share through both pricing and quality positioning. This competitive landscape aligns closely with the core assumptions of the Hotelling model, which features both differentiated products and multi-dimensional strategic competition. Given that pharmaceuticals are special commodities with significant product differentiation. Thus, we reinterpreted the physical space metaphor as a quality preference spectrum in the Hotelling model. Specifically, consumers are assumed to be uniformly distributed along a quality sensitivity interval [0,1], with firm A (low-tech) and firm B (high-tech) positioned at the two endpoints. A two-stage game-theoretic model is constructed based on this foundation. Following the implementation of NVBP policy, drugs were standardized through consistency evaluation and homogeneous grouping. With competition reduced solely to pricing, and the winning bidders selected based on the lowest price, the market structure now aligns with the core tenet of the Bertrand model, which is homogeneous products and pure price competition.

Model assumptions

Assumption 1

There are two pharmaceutical firms in the market,. Figure 1 shows the drugs they produce are located at 0 and 1 respectively, and the qualities are and.

Fig. 1.

Distribution of consumers and pharmaceutical firms

Assumption 2

From the perspective of pharmaceutical manufacturers, drug production must adhere to the Chinese Good Manufacturing Practice (CGMP) regulations. These regulations set the minimum quality standards for enterprises, and establishing a mature quality management system. The system requires significant investment in human resources, material resources, and financial costs, particularly in production technology. Therefore, we assumed that drug quality is an increasing function of the technical production costs incurred by the firm. The higher the technical investment a pharmaceutical firm makes, the higher the resulting drug quality. To facilitate analysis, we assume that the unit technical cost required to produce a drug with quality level is . reflects the technological cost of the pharmaceutical firm. A lower indicates a higher technological level, meaning the firm can produce higher quality drugs at lower technical costs. Without loss of generality, we assume indicating that the technological level of firm B is higher than that of firm A.

Assumption 3

Consumers are distributed between 0 and 1. Consumer utility is affected by four aspects: the distance \:x from the pharmaceutical firm to itself (reflecting consumer preferences), migration cost , drug quality, and drug price.

Parameter settings

Table 1 lists these parameters and their implications.

Table 1.

Parameters of the model

Parameters	Description
n	Number of firms
	Location of firm i on [0,1]
	Quality of firm i
	Cost coefficient (lower = higher technology)
	Cost function component, increasing in
	Marginal cost =
	Price of firm i
t	Transportation cost per unit distance
x	Consumer location
	Utility from firm i
	Profit of firm i
	Market share of firm i
	Indifferent consumer between i and i + 1
f(x)	Density of consumer distribution (uniform = 1)

Analytical procedure

Backward induction is the standard solution method for obtaining the subgame perfect Nash equilibrium in multi-stage games. In this study, we first solve the stage-two price competition: given any quality pair, we derive the best-response functions by solving the first-order conditions for profit maximization, yielding the equilibrium price expressions. Subsequently, we substitute these price equilibria into the stage-one profit functions, making profits dependent solely on quality choices. The equilibrium qualities are then determined by applying the first-order conditions for quality. This backward-solving procedure ensures that the resulting strategy profile constitutes a Nash equilibrium in every subgame, thereby eliminating incredible threats.

Results

Before NVBP policy implementation: two-stage game-theoretic model

Firms A and B compete for differentiation and can determine the quality of the drugs they produce. The utility of a consumer located at \:x (\:0<x<1) to purchase the two types of drugs is as follows:

1

2

The pharmaceutical firms engage in a two-stage game. In Stage I, they determine the quality of the drug to produce. In Stage II, they determine the pricing for the produced drug.

According to the backward induction method, as consumers in the drug market are usually single-home with the pharmaceutical firm, they only purchase a certain type of drug of a certain brand. The location of the indifferent consumer () who purchases drugs from the two pharmaceutical firms is:

3

The profit of the pharmaceutical firm is:

4

5

According to the backward induction method, we solve the first-order equilibrium conditions in Stage II, , , simultaneously to obtain the equilibrium quality and equilibrium price:

6

7

Proposition 1

From the above, when ,

8

9

10

Proposition 1 states that when pharmaceutical firms produce drugs of identical quality (), the technologically superior firm sets a lower price and captures a larger market share.The price difference is entirely attributable to the technical parameter: firm A (higher ) prices higher at , while firm B (lower ) prices lower at Consequently, the lower-priced firm B gains market share beyond the midpoint ().

Firms determine the pricing of their manufactured pharmaceuticals. We substitute the relationship between the drug price and quality into the profit function and solve the first-order equilibrium condition. Let. We solve them simultaneously to obtain the equilibrium quality and equilibrium price:

11

12

13

14

Proposition 2

can be obtained:

Proposition 2 shows two aspects of equilibrium behavior. First, the equilibrium quality of drugs produced by a pharmaceutical firm is inversely related to its own cost parameter and independent of competitor’s cost parameters. When a firm possesses superior technology, it tends to produce higher-quality drugs. Second, the equilibrium drug price is positively related to the firm’s own cost parameter but inversely related to competitors’ cost parameters(). A firm with inferior technology(higher ) tends to set higher prices to cover costs, but when facing competitors with superior technology, it lowers prices to attract consumers.

After NVBP policy implementation: price competition model

We assume which means only firm B has a higher technical level and can produce drugs of a certain quality. When the government implements centralized bidding and procurement, only one firm wins a bid in the market. China’s centralized drug bidding and procurement system has adopted a double-envelope system. The drugs are evaluated for consistency and the drug with the lowest price is required to win the bid after meeting the minimum quality requirements. we assume quality standard is the drug based on the existing drug quality in the market. When two pharmaceutical firms jointly participate in centralized bidding and procurement, because only one firm wins the bid at the lowest price, the profit function of the firm becomes

15

16

Case 1: Low quality standard

When both pharmaceutical firms are capable of producing drugs meeting quality standards and engage in Bertrand price competition, they set sufficiently low prices. Consequently, their equilibrium quality levels converge to the identical level . Firm B (with lower ) has marginal cost , while firm A(with higher ) has marginal cost .In a pure price competition (Bertrand) duopoly with homogeneous products, the firm with the lower marginal cost (Firm B) can undercut its rival and capture the entire market. The winning bid price is set slightly below Firm A’s cost, driving Firm A out of the market. The winning price is

17

18

Proposition 3

When , the winning bid price is , is the pre-NVBP equilibrium price of the same-quality drug produced by firm A, expressed as:

19

Specifically, if and only if . This condition is satisfied under Assumption 3.

Proposition 3 demonstrates that the procurement policy effectively reduces drug prices, but the post-procurement quality may be lower than the post-procurement market’s highest quality .

Case 2: High quality standard

When firm A fails the consistency evaluation, that is, , it is eliminated during the quality qualification stage and cannot produce compliant drugs. With only firm B qualifying for bidding, firm B’s monopoly pricing is set at:

20

Where the demand function depends on the government reserve price. If , .

Proposition 4

When , .

Proposition 4 demonstrates that when firm A fails the consistency evaluation, firm B can command monopoly pricing power. In this scenario, pharmaceutical firm lack competitive constraints, causing drug prices to instead become higher.

After NVBP policy implementation: post-award production decision

To analyze the winning firm’s post-award behavior, we introduce the following parameters: fixed production cost F(independent of output, assumed equal for both firms), default penalty cost (fines and blacklisting losses for failing to supply after winning), expected quality violation penalty (where is the probability of detecting quality deficiencies and is the fine amount), and the reduced quality level (where<). Given the winning price , the winning firm decides whether to fulfill the contract and whether to produce at the standard quality level, under the principle of profit maximization.

After winning the bid, firm B must decide whether to fulfill the contract. If it produces at the minimum standard and fulfills the contract, its profit is:

21

If the firm chooses to default (not fulfill), its profit is Default occurs when the profit from compliance is less than the profit from default:

22

Define the default threshold price:

23

When , the firm chooses to default. The lower the winning price, the higher the default risk.If the firm chooses to fulfill, it must also decide on the quality level. If it reduces quality from the standard to (where< it saves production cost but faces an expected penalty . The firm reduces quality if:

24

i.e. “the cost saving exceed the expected penalty.” Define the quality risk threshold price:

25

When , the firm’s profit margin is insufficient to cover high-quality investment, creating an incentive to reduce quality. The lower the winning price, the higher the quality risk. Comparing the default threshold and the quality risk threshold , the winning price can be divided into three intervals, including no risk, quality risk and supply disruption (Table 2).

Table 2.

Firm behavior and risk types under different winning price intervals

Price interval	Firm behavior	Risk type
	Fulfills, maintains standard quality	No risk
	Fulfills but may reduce quality	Quality risk
	Defaults	Supply disruption risk

Proposition 5

The winning price may fall below the firm’s quality incentive threshold or even the default threshold . When , the firm has an incentive to reduce quality (quality risk); when the firm chooses to default (supply disruption risk). Greater price reductions increase both types of risks, but the two risks are inversely related: when prices are so low that default occurs, the quality risk no longer exists (because production has ceased). Policy design should keep prices within the safe interval [ to avoid triggering either quality risk or supply disruption risk through excessive price compression.

Robustness analysis: extension to n firms

To assess the robustness of our main findings (Propositions 1 and 2), we extend the baseline duopoly model to an n-firm setting. Extending the model to n firms serves to verify whether the duopoly conclusions remain robust under more general and realistic market structures, characterized by multiple firms, thereby enhancing the theoretical model’s external validity and bolstering the credibility of its policy implications. All assumptions from section “Methods” remain unchanged, except:

• There are chemical generic drug manufacturers.

• firms are located non-uniformly at positions with .

• Consumers are uniformly distributed on with density .

• The utility of a consumer at location x buying from firm i is.

26

• Consumers choose the firm offering the maximum utility to purchase drugs.firm i’s marginal cost is ,with as defined before.

The game remains two-stage: firms choose qualities first, then prices. We solve by backward induction.

Stage II: Price competition

Given the quality vector , firms choose the price vector to maximize profits. Market share depends on the prices and qualities of all firms. The location of the indifferent consumer between firm \:i and its adjacent firm (e.g., firm and firm ) is determined by the utility equality condition. For example, the indifferent point between firm \:i and firm \:j satisfies:

27

Assuming ordered locations (, for firm \:i and firm \:i+1, within the interval , the indifferent point is:

28

profit function is:

29

where is the market share of firm (i.e., the proportion of consumers choosing firm \:i).First, we apply backward induction to derive the second-stage price competition:

The market share of firm \:i is:

30

where the boundary conditions are: (specify for the leftmost firm i = 1,define =0;for the rightmost \:i=n, define ), the first-order condition for profit maximization is:

31

Solving this yields:

32

This equation defines the price equilibrium, but generally lacks an analytical solution and requires numerical solving. Under the assumption of symmetric firm locations(i.e.,firms are equidistant from each other, so ) and symmetric quality levels( for all i), but allowing for heterogeneous technology parameters , the equilibrium price for firm i is:

33

This shows that even with identical locations and quality, technology differences persist in equilibrium pricing.

Stage I: Quality choice

Given the price equilibrium, we proceed to derive the first-stage quality competition. Under general locations and distributions, this condition becomes analytically complex. However, in the symmetric case (where all , locations are equidistant, and consumers are uniformly distributed), the equilibrium quality satisfies:

34

The n-firm analysis is moved to the Appendix A as a robustness check.

Proposition 5

(Price difference with equal quality, n firms): When all firms produce the same quality \:q, equilibrium prices are driven by both technology parameters and locations. For firm i and i + 1,the price difference satisfies:

35

Thus, a technologically superior firm (lower ) may set a lower price if location differences are not too large. This generalizes Proposition 1.

Proposition 6

(Comparative statics with n firms):.Under symmetric firm locations (equidistant) and symmetric , the equilibrium quality satisfies identical to the duopoly case. In asymmetric locations, a firm’s quality choice is affected only by its own and the identity of its immediate neighbors; the condition still holds, while cross-effects ( for non-neighbors) are zero. For prices, (higher → higher price) and for j being a neighbor, the same signs as in Proposition 2, but the effect decays with distance.

Discussion

This study focuses on the relationship between drug quality and price before and after the implementation of the NVBP policy. Using a two-stage game (duopoly quality-price competition) and a double-envelope bidding game (simulating centralized procurement), we reveal that after the implementation of the NVBP policy, there may be risks of reduced drug quality and market monopolization. The following discussion elaborates on these findings, clearly distinguishing between results directly derived from our model and those supported by existing empirical studies.

Key model findings

• Before NVBP policy implementation: Proposition 1 suggests that when firms produce the same quality, the technologically superior firm sets a lower price and captures a larger market share. Proposition 2 indicates that when firms endogenously choose quality, a technologically superior firm produces higher quality and sets a higher price. Thus, the model indicates that in the absence of procurement policy, a market mechanism exists where quality improvement can potentially be rewarded through price and market share advantages.

• After NVBP policy implementation: Proposition 3 and 4 suggest two cases of quality and price competition. Under a low quality standard, firms compete aggressively on price, leading to a new equilibrium with lower quality investment relative to the pre-NBVP policy level. Under a high quality standard that only the technologically advanced firm can meet, the market may result in a sole-supplier situation, where the winning firm gains exclusive supply rights. Proposition 5 suggests the quality risk and supply disruption risk may occur when the price of drugs is too low.

Firm behavior before NVBP policy implementation

Multiple empirical studies support the result that before the NVBP policy, China’s pharmaceutical market was highly competitive, and firms competed through quality improvement and innovation. Technologically advanced firms, through substantial R&D investments, indeed develop superior drug formulations and more efficient manufacturing processes, resulting in higher-quality drugs at lower unit costs [30]. This pattern is broadly consistent with the insight that superior technology leads to higher quality. Moreover, the market rewarded such quality differences: higher-quality drugs commanded higher prices, and patients/doctors showed preference for better products, a pattern that consist with the positive relationship between quality and price in Proposition 2.

Proposition 1 predicts that if two firms happen to produce identical quality, the technologically superior firm will charge a lower price and gain more market share. Empirical evidence on this specific scenario is scarce, because in reality firms rarely produce exactly the same quality; quality differentiation is the norm. However, a substantial body of research on generic drug markets suggests that the establishment of therapeutic equivalence through regulatory pathways (i.e., quality assurance) is a critical prerequisite for enabling price competition, under which more efficient manufacturers are expected to capture market share by offering lower prices [31]. Thus, while Proposition 1 offers a plausible baseline prediction, its core assumption of perfect quality disclosure may not always hold. For instance, winning bidders sometimes hide quality deficiencies which could theoretically blur the predicted price-market share relationship [17]. However, the NVBP policy provides a unique quasi-experimental setting to test the boundaries of this proposition. This suggests that the predicted relationship is resilient and that future research could fruitfully extend the model by incorporating quality disclosure as an endogenous strategic variable.

Firm behavior after NVBP policy implementation

Existing studies confirm that the NVBP policy has significantly reduced drug prices and compressed profit margins in the short term [32, 34]. Our model offers a potential theoretical explanation for this pattern: a low quality standard shifts competition from quality to price. Furthermore, data analysis suggests that NVBP policy led to a 40% drop in R&D investment for the first batch of varieties (2019–2022) [10], indicating weakened incentives for quality improvement, which broadly consistent with the model’s prediction of reduced quality investment under a low standard. Our model further predicts that quality risk and supply disruption risk may emerge after the policy implementation. Illustrative cases that resonate with these theoretical risks exist in reality. For example, although Methocarbamol Injection (manufactured by Chengdu Tiantai Mountain Pharmaceutical Co., Ltd. on behalf of Guangzhou Hehe Pharmaceutical Co., Ltd) was selected in the 11th batch of China’s NVBP policy, subsequent special inspections identified serious quality deficiencies. The Joint Procurement Office consequently revoked its winning status and blacklisted both firms, suspending their NVBP participation until October 2027 [35]. Table 3 compiles additional post-NVBP incidents involving quality or compliance deficiencies across multiple batches. We caution that these cases are illustrative rather than systematic empirical evidence of a causal link between the NVBP policy and quality degradation. The observed deficiencies may reflect pre-existing regulatory gaps, firm-specific governance failures, or broader industry challenges rather than solely post-award quality dilution. Consequently, we present these cases as qualitative illustrations that resonate with our model’s predictions, rather than as definitive proof of the model’s external validity.

Table 3.

Risk events after NVBP policy implementation

Drug Name	Batch	Enterprise	Quality Incident
Paclitaxel for Injection (Albumin-bound)	2nd Batch	Celgene Corporation	Mar 2020: Disqualified, Inadequate Sterility Control
Glimepiride Tablets	2nd Batch	Guizhou Shengjitang Pharmaceutical	Jan 2022: Disqualified, Serious GMP Deficiency
Dutasteride Capsules	5th Batch	GSK	Nov 2022: Disqualified, Not Tested per Registration Standards
Bicalutamide Tablets	5th Batch	Sun Pharmaceutical (India)	Jun 2022: Disqualified, Loss on Drying Failed
Cefodizime Sodium for Injection	8th Batch	Daewoong Bio (Korea)	Jul 2024: Import Suspended, Refused Overseas Inspection
Atomoxetine Hydrochloride Capsules	9th Batch	Dr. Reddy’s	Aug 2024: Import Suspended, API Management Deficiency
Phloroglucinol Injection	10th Batch	Sichuan Haimeng Zhisen / Taiji Pharmaceutical	Mar 2025: Disqualified, Serious GMP Deficiency
Methocarbamol Injection	11th Batch	Guangzhou Hehe / Chengdu Tiantaishan	Apr 2026: Disqualified, Serious Deficiency
Piracetam Injection	11th Batch	Zhuhai Hefan / Shandong Weigao	Apr 2026: Disqualified, Serious Production Management Deficiency

Furthermore, our model predicts an extreme scenario: if the government sets drug quality standards so stringent that only the technologically most advanced firm can meet them, the market could theoretically evolve toward a monopoly. Direct empirical evidence for this extreme case under chemical generic NVBP is scarce, precisely because policymakers typically refrain from setting quality standards excessively high in practice, so as to avoid creating a de facto monopoly. Nevertheless, loosely analogous dynamics have been reported in Traditional Chinese Medicine (TCM) patent medicine procurement, where the absence of robust competition among exclusive products has been associated with persistent pricing challenges [36].

Policy implication

Regarding the selection of reference preparations for consistency evaluation, authorities may wish to avoid setting the quality bar at a minimally sufficient level. Game-theoretic analysis within our stylized NVBP framework indicates that if the government establishes only a minimal quality threshold, pharmaceutical manufacturers may prioritize price competition over quality investment. Under such conditions, firms might lack the economic incentive to improve drug quality, because investments in technological innovation and process optimization could erode price competitiveness by increasing production costs rather than yielding returns in the bidding process. In scenarios where a single firm exclusively manufactures a drug, it may be prudent to strengthen both quality supervision and price regulation. Policy measures, such as intensified regulatory oversight and antitrust monitoring, could help mitigate the emergence of an exclusive monopoly.

Finally, policy design could explore incorporating quality indicators into the scoring system, thereby drawing on the intrinsic logic revealed in Proposition 2 to convert quality differentiation into quantifiable price advantages and restore a positive quality-price linkage.In summary, the following policy recommendations are proposed as theoretically grounded hypotheses:

• Quality standards should be set carefully to balance price competition and quality incentives, recognizing that excessively low standards may, according to our model, trigger price undercutting at the expense of quality.

• For sole-supplier pharmaceutical firms, enhanced quality supervision and price regulation during and after bidding may be warranted.

• Policymakers could explore incorporating more quality indicators into the scoring system to restore the positive quality-price link.

Conclusions

This study develops a game-theoretic model to examine how China’s NVBP policy reshapes pharmaceutical firms’ quality and pricing strategies. The analysis yields two main comparisons. Before the policy, the model indicates that firms with technological advantages have endogenous incentives to produce higher-quality drugs, and market mechanisms may reward such improvements through price and market share advantages. After the policy, the competitive dynamic shifts: a low quality standard might trigger pure price competition and may reduce quality investment, while a high standard that only the technologically advanced firm can meet may create a sole-supplier situation with monopoly pricing power. Additionally, excessively low winning bid prices may generate quality or supply disruption risks.

These findings suggest that policy calibration matters. To mitigate unintended consequences, tiered incentive mechanisms that restore a positive link between quality and pricing, alongside strengthened post-market surveillance, appear important. The model provides theoretical insights for balancing cost control with quality assurance, though empirical validation remains a necessary next step.

Limitations

This study models the game-theoretic relationship between drug quality and price of two firms before and after the implementation of the NVBP policy. However, there still exist some limitations in our model. Firstly, while the model provides clear analytical tractability by assuming symmetric competition between two firms, this simplification may limit the direct generalizability of our findings to markets with highly asymmetric firm capabilities or non-uniform consumer preferences. Although our robustness analysis with n firms supports the stability of the main propositions under symmetric assumptions, the lack of a fully integrated analytical solution for general asymmetric cases means that the quantitative predictions (e.g., exact price differentials) should be interpreted as illustrative of directional effects rather than as precise empirical forecasts. Secondly, by focusing solely on manufacturer competition, our model abstracts away from the critical roles of insurers, hospitals, and patients. In reality, these stakeholders’ behaviors—such as formulary decisions, prescribing habits, and patient adherence—can significantly influence market outcomes. Omitting these demand-side feedback loops may cause our model to overestimate the price-reducing effect of competition or underestimate the quality degradation risk, as it does not account for the potential countervailing power or preferences of downstream actors. Thirdly, the study remains theoretical in nature. While we use real-world cases to illustrate the model’s qualitative predictions, we do not employ econometric estimation to test these predictions. This lack of empirical validation means that our findings should be viewed as hypothesis-generating rather than hypothesis-confirming. Future research could bridge this gap by using disaggregated procurement data or designing quasi-experimental studies to empirically test the proposed relationships between quality standards, winning bids, and post-award firm behavior.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Liang Zhu is the lead corresponding author, responsible for primary correspondence with the journal,including manuscript submission, revision and respnse to peer review comments.Changsheng Chen is the co-corresponding author, who assisted in coordinatin academic communications and verifying specific techincal details of the study. Rui Wang and Jun Li searched for literature and revised the background section. Jiaxue Li and Siming Li analyzed and interpreted the data. Siyu Zhang verified the data results, drew the graphs, and drafted the manuscript. Linjing Zhang revised the discussion section. All authors reviewed the manuscript.

Funding

This work was supported by National Natural Science Foundation of China, Named Research on the negative unintended effects of centralized volume-based drug procurement policy from the perspective of dynamic game: identification, verification, formation mechanism and policy optimization (Grant:72474221), and it was also supported by Young Talents Support Program(originates from School of Public Health, The Fourth Military Medical University), named Research on the Collaborative Governance Mechanism for Primary-Level Drug Supply Assurance under the Framework of County-Level Medical Consortiums.

Data availability

Not applicable.

Declarations

Ethical approval

Not applicable.

Disclaimers

Not applicable.

Competing interests

The authors declare no competing interests.