Abstract
Firms typically expand R&D in response to underperformance, yet research has treated this response as firm-centric and overlooked how national institutions shape its intensity. Building on the Behavioural Theory of the Firm and literature on government and institutions, we develop a cross-level framework in which two government levers, direct R&D support and digital capacity, shape the extent to which firms intensify R&D after underperforming. Using a 10-year panel of manufacturing firms, we find that underperformance is positively associated with R&D intensity and that both government R&D support and government digital capacity strengthen this relationship. We further hypothesise and empirically test that the moderating effect of government R&D support is particularly stronger where government digital capacity is high. These results are robust to multiple tests addressing econometric specifications and temporal coverage. Theoretical and practical contributions and avenues for future research are discussed.
Similar content being viewed by others
1 Introduction
Research on the Behavioural Theory of the Firm (BTOF) has long examined how firms adjust their R&D intensity in response to performance feedback (Cyert & March, 1963; Gao et al., 2023; Hu et al., 2022; Zhang et al., 2024). A central tenet of the BTOF is that decision-makers evaluate outcomes relative to aspiration levels, interpreting them as either underperformance or overperformance (Greve, 1998; Ref & Shapira, 2017; Rhee et al., 2019; Yu et al., 2019). Classic studies demonstrate that underperformance relative to aspirations triggers problemistic search, i.e., a search for solutions prompted by a problem, leading firms to intensify R&D investments (Cyert & March, 1963; Greve, 2003; Kuusela et al., 2017; Levinthal & March, 1981; Xu et al., 2019). More recent research, however, emphasises that this relationship is not uniform but shaped by important firm-level boundary conditions such as financial vulnerability (Chen & Miller, 2007), managerial overconfidence (Zhao et al., 2025), strategic orientation (Kotiloglu et al., 2021), the persistence of underperformance (Yu et al., 2019), prior experience (Huang et al., 2021), and business strategy (Madadian & Van den Broeke, 2023).
While this stream of research has provided valuable insights, its predominantly firm-centric lens has left the role of the broader institutional context undertheorised (Gao et al., 2023; Han, 2023). This is particularly important because firms are not isolated actors but are embedded in national institutional environments that set the “rules of the game” and shape access to external resources and information (Chen & Miller, 2007; Goyal, 2025; Peng, 2003). When R&D investments demand substantial resources (Watanabe et al., 2002) that underperforming firms struggle to mobilise internally, the supporting and/or constraining institutional context becomes especially consequential. In this regard, a long-standing stream of research emphasises the role of government, showing that government policies and infrastructure help shape the institutional environment in which firms operate (Gao et al., 2023; Hillman & Keim, 1995; Nave et al., 2024; Rothstein & Teorell, 2008). Published studies show that the government-driven institutional arrangements can function as either supportive mechanisms (Doh & Kim, 2014; Micol et al., 2025) or constraining forces (Guan & Yam, 2015), ultimately influencing firms’ strategic decisions. Yet, despite these insights, the BTOF literature has remained largely silent on this dimension.
In this paper, we build a bridge between the BTOF’s micro-level performance feedback models and macro-level government policies and infrastructure. We argue that to understand why firms’ R&D response to underperformance varies systematically across contexts, it is necessary to look beyond firm-level factors and consider the government-level conditions that structure those contexts. In particular, we focus on two dimensions. First, government R&D support, particularly direct funding programs (Kang & Park, 2012; Lenihan et al., 2024), which can ease financing constraints, share risk, signal credibility, and broaden search. Second, government digital capacity which refers to how effectively governments use digital channels to make program information and access interfaces visible, transparent, and usable for firms (Malodia et al., 2021; Williams, 2023). Together, we argue that these government levers define key contextual features that influence how performance feedback is translated into problemistic search and R&D intensity.
We thus build on these considerations and ask: How do government R&D support and government digital capacity shape firms’ R&D response to underperformance? We further theorise that these two factors are not only independently significant but also complementary. The effectiveness of government R&D support ultimately depends on whether firms can identify and access available programs in a timely and transparent manner, a process facilitated by high digital capacity. Accordingly, we also test a three-way interaction between underperformance, government R&D support, and government digital capacity to capture this synergistic effect. Drawing on a 10-year panel dataset of manufacturing firms across 30 countries, our analysis first confirms the BTOF baseline: underperformance is positively associated with R&D intensity. We then find that both government R&D support and government digital capacity independently accentuate this positive relationship. Importantly, our results support the three-way interaction, indicating that the moderating effect of government R&D support is stronger in institutional environments with high government digital capacity.
With this study, we make two important contributions. First, we advance the BTOF by demonstrating that the intensity of problemistic search is shaped not only by aspiration–performance dynamics and firm-internal contingencies, but also by government-level factors that condition how strongly firms can translate underperformance into R&D investment. In particular, we show that both government R&D support and government digital capacity increase firms’ R&D response to underperformance, thereby shifting the conversation from the well-established question of whether underperformance triggers R&D to the more nuanced question of how strongly this trigger translates into R&D resource commitments under varying national conditions. This focus complements the other work on institutional quality (North, 1990) and quality of government (Rothstein & Teorell, 2008) which primarily explains cross-national differences in baseline feasibility of market activity and innovation by theorizing and testing how two specific, policy-relevant levers shape the responsiveness of firms’ problemistic search to performance feedback. Second, we develop a cross-level perspective (Klein et al., 1994) in which firm-level performance feedback operates within country-level policy and information regimes. We also demonstrate that government R&D support and digital capacity are complementary: while funding expands the feasible set of R&D responses, digital capacity shapes the visibility, timeliness, and ease with which firms can identify and access those programs and relevant knowledge. This configuration logic explains systematic cross-national variation in the magnitude of underperformance-induced R&D (Xue et al., 2023), extends contingency thinking beyond firm-internal or cognitive moderators, and offers a pathway for reconciling inconsistencies in prior work that documents context dependence and fragmented findings in performance feedback effects on R&D (Martínez-Noya & Valdés-Llaneza, 2025). More broadly, by testing a three-way interaction, we connect multilayer policy design to managerial decision-making, explaining how governments influence both firms’ option sets and the frictions they face when adapting under performance pressure.
2 Theoretical background
The BTOF explains organisational strategic decisions as an aspiration-driven, feedback-based process under bounded rationality: because decision makers face cognitive and informational limits, they evaluate outcomes relative to a salient aspiration level rather than through exhaustive optimisation (Cyert & March, 1963; Gao et al., 2023; Gavetti et al., 2012). Aspirations are set using backward-looking and social reference points, i.e., own past performance and that of comparable peers, so that a continuous performance metric is categorised as “success” (above aspirations) or “failure” (below aspirations) for purposes of action (Baum & Dahlin, 2007; Kuusela et al., 2017; Dong et al., 2021). Performance feedback is thus the perceived discrepancy between the firm’s actual performance and aspirations, and it operates as a heuristic trigger for organisational responses (Argote & Greve, 2007; Greve & Gaba, 2017). When feedback indicates underperformance (below aspiration—failure), the BTOF predicts problemistic search and the intensity of this search increases with the size of the aspiration shortfall (Baum et al., 2005; Cyert & March, 1963; Greve, 2003). This mechanism has been used to account for a wide range of strategic choices such as bribery (Xu et al., 2019), alliance formation (Han, 2023), reshoring activities (Zhang et al., 2024), foreign exit decisions (Dai et al., 2023), and internal governance changes (Sengul & Obloj, 2017).
One particular response to underperformance is R&D investment, a widely observed form of problemistic search because R&D, thereby innovation, offers a direct path to performance recovery (Cyert & March, 1963; Gavetti et al., 2012; Posen et al., 2018). A large empirical literature shows that firms tend to increase R&D intensity when they underperform (e.g., Greve, 2003; Chen, 2008; Lim & McCann, 2014; O’Brien & David, 2014; Rhee et al., 2019; Rudy & Johnson, 2016; Carnes et al., 2022; Gao et al., 2023; Zhang et al., 2024). The core mechanism is that R&D enriches the firm’s resource base and enables new products, services, and processes that reposition the firm competitively (Lim & McCann, 2014; O’Brien & David, 2014). Additionally, because innovative capabilities are path dependent and hard to imitate, intensified R&D can yield durable rents, further justifying investment under underperformance (Barney, 1991; Greve, 2003). Consistent with this logic, decision makers frequently read underperformance as innovation deficiencies and step up R&D to reverse decline (Ehls et al., 2020; Gao et al., 2023; Huang et al., 2021; Lantz & Sahut, 2005; Scoresby et al., 2021; Zhang et al., 2024). Moreover, extant literature has shown that the magnitude of the R&D response scales with the size of the performance gap, as larger underperformance increases urgency, risk tolerance, and exploratory search (Iyer & Miller, 2008).
More recently, research has shown that the R&D response to underperformance is not uniform and varies with firm-level contingencies. For example, Chen and Miller (2007) show that slack resources enable, whereas proximity to bankruptcy constrains, R&D search intensity. Yu et al. (2019) add that the duration of underperformance matters: as underperformance persists, firms first retrench and then re-intensify and broaden R&D search, yielding nonlinear effects. Madadian and Van den Broeke (2023) further demonstrate that a firm’s risk profile and business strategy shape responses, with high-risk firms tending to increase R&D when performance declines, whereas low-risk firms maintain or reduce it. Finally, a meta-analytic review synthesizing 113 studies shows that internal attributes such as firm size, age, ownership, and slack systematically moderate the underperformance–R&D link (see Kotiloglu et al., 2021 for a review).
However, scholars have noted that the contingencies studied remain predominantly firm-centred, leaving the institutional context undertheorised (Gao et al., 2023; Han, 2023). Yet, firms are embedded in national environments that set the “rules of the game” and shape access to resources and information (Chen & Miller, 2007; Gao et al., 2023; Goyal, 2025; Han, 2023; Peng, 2003). In the R&D domain, overlooking national context is particularly consequential: R&D is resource-intensive (Watanabe et al., 2002) and underperforming firms often lack sufficient internal resources, making supportive and/or constraining institutions pivotal. Additionally, and more broadly, theoretical advancement hinges on clarifying external boundary conditions, i.e., specifying when and where core mechanisms are expected to change, to ensure properly scoped and generalisable claims (Bacharach, 1989; Busse et al., 2017; Edwards & Berry, 2010; Whetten, 1989). Building on this view, recent work on the BTOF and R&D urges a shift from internal, behavioural moderators toward external institutional influences (Gao et al., 2023; Gavetti et al., 2012; Han, 2023). In this regard, a long-standing body of research shows that the government, through policy programmes and infrastructure investment, acts as the primary architect of the institutional environment that shapes firms’ strategic choices, ultimately enabling and constraining organisational responses (Doh & Kim, 2014; Guan & Yam, 2015). This perspective positions government as a pivotal institutional boundary condition (Gao et al., 2023; Hillman & Keim, 1995; Ozturk-Kose, 2025; Rothstein & Teorell, 2008) that can help explain how R&D intensity as a response to underperformance changes and warrants explicit theorisation.
We thus focus on two government-driven institutional levers shown to be especially salient in shaping firm strategic behaviour: government R&D support (Doh & Kim, 2014; Kang & Park, 2012) and government digital capacity (Malodia et al., 2021; Williams, 2023; Xu & Jin, 2024). In the following sections, we develop the arguments for each lever and then theorise their complementarity in shaping firms’ R&D response to underperformance.
3 Hypotheses development
3.1 Underperformance and R&D intensity
We first establish the baseline hypothesis. In accordance with the established BTOF research, when firms underperform relative to their aspiration level, they initiate problemistic search through which they increase their R&D. In fact, R&D is especially important in this context because it expands the set of recovery options and builds capabilities that can reposition the firm competitively (Lim & McCann, 2014; O’Brien & David, 2014). In this regard, several studies across industries and countries consistently and empirically show that firms below aspirations devote a larger share of revenue to R&D than peers at or above aspirations, and that the response strengthens as the gap below the aspiration level grows, manifesting in both higher R&D intensity and a preference toward more exploratory efforts (Bromiley & Washburn, 2011; Carnes et al., 2022; Chen, 2008; Falk, 2012; Gao et al., 2023; Greve, 2003; Iyer & Miller, 2008; Rhee et al., 2019; Rudy & Johnson, 2016; Zhang et al., 2024). Thus, in line with the BTOF, we also hypothesise:
Hypothesis 0 (Baseline)
Underperformance is positively associated with a firm’s R&D intensity.
3.2 The role of government R&D support
While the baseline hypothesis is that underperformance raises R&D intensity, a firm’s ability to act on this pressure depends on how supportive its institutional environment is for initiating and sustaining R&D-oriented problemistic search (Dimos & Pugh, 2016; Huang et al., 2021). In this regard, we conceptualise government R&D support, particularly direct funding, as a country-level institutional moderator that conditions how strongly underperformance translates into R&D intensity by shaping the feasibility and attractiveness of increasing R&D. Thus, we argue government R&D support provides an enabling institutional basis for problemistic search, thereby positively moderating the underperformance–R&D intensity relationship.
Implicit in this view, there are multiple key reasons. First, direct government R&D support, as an institutional instrument, injects capital that creates or supplements organisational resources, making it feasible for underperforming firms to sustain R&D without jeopardizing core operations (Xu et al., 2021). Second, government R&D support lowers the effective cost of experimentation and shares downside risk, reducing decision makers’ perceived likelihood of failure (Guellec, 2003; Jia et al., 2021). These effects are especially salient when underperformance is large and urgency to recover is high, encouraging high-risk, high-return innovation options aimed at reversing decline (Pearce et al., 2011). Third, the award of a government grant operates as a quality signal that can unlock complementary private financing and investor support, expanding the feasible set of R&D responses (David et al., 2000; Dimos & Pugh, 2016). Beyond resource provision, support can also confer legitimacy on internal strategic choices, reinforcing managerial commitment and organisational alignment around R&D (Zhou et al., 2024). Moreover, the influence of government R&D support is not only financial. Research on behavioural additionality suggests that government R&D programs, as components of the institutional environment, encourage collaborative behaviours, widen search breadth, and increase openness to external knowledge (Buisseret et al., 1995; Clarysse et al., 2009; Cunningham et al., 2016), all of which contribute to more extensive problemistic search.
A plausible counter-mechanism is resource-induced satisficing: readily available funds may dampen search intensity by easing short-term pressure. In our context, two features mitigate this risk. First, underperformance heightens threat perceptions and recovery imperatives, making resources more likely to be deployed toward exploration rather than complacency (Audia & Greve, 2006). Second, direct R&D funding is typically earmarked and milestone-based, limiting diversion and encouraging continued experimentation (Guo et al., 2016). Accordingly, the enabling and discipline effects of government R&D support should outweigh satisficing tendencies among underperforming firms.
Based on all of the above considerations, we thus advance the following hypothesis:
Hypothesis 1
The positive relationship between underperformance and R&D intensity is stronger (weaker) when government R&D support is higher (lower).
3.3 The role of government digital capacity
Government digital capacity captures how effectively governments use digital channels to make program information and access interfaces visible, transparent, and usable for firms (Malodia et al., 2021; Williams, 2023). In our framework, digital capacity represents a country-level institutional condition that shapes the information environment in which underperforming firms search for and pursue R&D options. Accordingly, we expect government digital capacity to positively moderate the relationship between underperformance and R&D intensity.
In digitally capable institutional contexts, where procedures are clear and consistent and information is accessible, underperforming firms can more readily discover relevant R&D programs and information, assess eligibility and timelines with confidence, and pursue applications with greater certainty (Guo et al., 2022; Petrin, 2018; Stephan et al., 2015). Timely, searchable, and standardised disclosures about supporting program rules, deadlines, and award histories, alongside open data resources, reduce discovery costs and accelerate the mobilisation of information for R&D options (Li & Xu, 2024). This is particularly important for problemistic search because it reduces interpretive ambiguity and lowers the expected cost of identifying feasible responses. In fact, government digital capacity improves the transparency and usability of the information environment in which underperforming firms search for recovery through R&D options.
Digital capacity also affects whether information can be converted into access. Where governments provide predictable online procedures and integrated workflows, firms can translate awareness of opportunities into applications and decisions with fewer delays, inconsistencies, and compliance burdens (Chen et al., 2024). From a BTOF perspective, this reduces the perceived risk of wasted effort, a central deterrent to search when firms face time pressure and high uncertainty under underperformance. Accordingly, digital capacity reduces both information frictions in locating and interpreting opportunities and access frictions in acting on them, which together condition the intensity of underperformance-driven R&D responses.
Additionally, high government digital capacity enhances benchmarking and aspiration formation by making peer and market data more immediate, standardised, and widely accessible (OECD, 2020), thereby sharpening performance feedback and focusing problemistic search on higher-quality opportunities. In other words, digital capacity improves the visibility and comparability of performance relevant information that firms and stakeholders already use, reducing noise in interpreting feedback and supporting more directed search. In fact, government digital capacity shapes an institutional environment in which access to external knowledge through open data, digital research repositories, and e-procurement is easier and faster, thereby, lowering search and contracting costs and expanding feasible search trajectories without proportionate cash outlays (Chen et al., 2024; Liu et al., 2024). In this way, government digital capacity functions as an informational and access infrastructure that expands the set of actionable options available to firms responding to performance shortfalls.
We acknowledge potential countervailing forces, such as digital divides or usability hurdles, but note that mature digital systems emphasise interoperability, user-centric design, and support services, which dampen these risks (Liu et al., 2024). On balance, we predict that government digital capacity strengthens the underperformance and R&D intensity link by reducing information and access frictions and lowering uncertainty, thereby enabling underperforming firms to translate performance shortfalls into more intensive problemistic search and higher R&D intensity.
Hypothesis 2
The positive relationship between underperformance and R&D intensity is stronger (weaker) when government digital capacity is higher (lower).
3.4 The impact of government digital capacity on the role of government R&D support
Hypothesis 1 predicts a stronger positive relationship between underperformance and R&D intensity when government R&D support is higher. We now refine this hypothesis and argue that the magnitude of this moderating effect is itself conditional on government digital capacity. In our theorisation, R&D support functions as an institutional moderator that relaxes financial constraints for underperforming firms, but its behavioural bite depends on the institutional information regime shaped by digital capacity. In fact, high digital capacity renders government R&D support programs more visible and recognisable. Under underperformance, when attention and liquidity are scarce, this visibility makes government support more discoverable in time, eligibility legible, and application sequencing actionable. Thus, the same government support programs exert a stronger moderating influence in digitally capable environments because firms in need can more readily find, interpret, and mobilise it toward R&D. Additionally, government digital capacity improves the allocative quality of support programs by supplying reliable, timely information and a transparent, interoperable system that guides decision makers in underperforming firms to channel government support toward more ambitious, exploratory R&D, rather than defaulting to incremental, criteria-conforming projects. In contrast, where digital capacity is low, the same government support is harder to target and more likely to be misallocated; applications stall or are steered into low-variance efforts.
Accordingly, we argue that government R&D support and digital capacity operate as institutional complements: support expands the feasible R&D resource set, while digital capacity makes those resources visible and accessible and directs them toward high-impact uses, thereby increasing support’s moderating influence. We thus advance the following hypothesis:
Hypothesis 3
The positive moderating effect of government R&D support on the relationship between underperformance and R&D intensity is stronger when government digital capacity is higher.
4 Method
4.1 Sample
Firm-level data for this study were obtained from Bureau van Dijk’s Orbis database, a widely recognised source of firm-level data encompassing annual financial statements, ownership structures, and business activities (Bhaumik et al., 2025; Von Nitzsch et al., 2024). To construct the sample, we focused on manufacturing firms that are typically among the most R&D-intensive sectors (Bae et al., 2008). We used the NACE industry classification system to identify manufacturing firms and select those classified under codes 10 to 33. This broad classification includes a wide range of manufacturing activities (such as food products, textiles, pharmaceutical products, basic and fabricated metals, and electrical and machinery equipment, etc.) and offers a representative sample. To construct the sample, we developed a ten-year panel covering the period from 2011 to 2020, focusing on firms’ R&D intensity. We limit the sample to up to 2020 to avoid the exogenous shock introduced by the COVID-19 pandemic given that this disrupted global markets and firm behaviour, particularly R&D intensity. From Orbis, we initially identified 232,326 manufacturing firms (NACE 10–33) with an active status and at least one employee in 2020 (the final year of our panel). We then applied a series of standard sample-construction screens to ensure the availability of the focal firm-level variables and a consistent match to country-level variables over time.
Government-level data were drawn from two primary sources. First, information on government R&D support was obtained from the OECD’s R&D Tax Expenditures and Direct Government Funding of BERD database. This source offers comprehensive indicators that capture the scale and composition of both central and subnational government support for business R&D across OECD countries and other major economies. It includes measures of direct government funding based on firm-level reports collected by national statistical agencies, providing a holistic view of government efforts to encourage R&D investment. Second, to assess government digital capacity at the country level, we merged our firm-level dataset with the United Nations E-Government Development Index (EGDI). The EGDI is a widely used composite index based on a biennial global survey of all 193 UN Member States to evaluate the capacity of governments to provide public services through digital platforms (e.g., Akpan-Obong et al., 2023; Dhaoui, 2022; Seiam & Salman, 2024). The index is composed of three equally weighted sub-indices: the Online Service Index (OSI), which evaluates the scope and effectiveness of government websites; the Human Capital Index (HCI), which captures education-related metrics such as adult literacy and school enrolment; and the Telecommunication Infrastructure Index (TII), which measures connectivity through indicators like internet penetration, mobile subscriptions, and broadband access. As the EGDI is published biennially, we applied linear interpolation, consistent with Dhaoui (2022), to estimate values for the intervening years, an approach justified by the relatively stable progression of government digital capacity metrics over time.
Additionally, we relied on the World Bank’s and World Intellectual Property Organization’s (WIPO) databases for several country-level variables. The World Bank database is a widely used and authoritative source that provides consistent, high-quality economic and demographic data across a broad set of countries. Its comprehensive coverage and standardisation make it particularly well-suited for cross-national research, especially when incorporating macro-level contextual factors.
From the 232,326 active manufacturing firms identified in Orbis, we first excluded firms with missing firm-level information required to compute R&D intensity and the firm-level variables used in the analyses, resulting in 26,736 firms. We then merged the firm-level sample with the country-level sample and excluded firms in countries for which the government R&D support, the government digital capacity, and other country-level variables were unavailable, yielding 17,415 firms. Finally, to construct a balanced ten-year panel for 2011–2020, we retained only firms with complete observations across all ten years and dropped the remainder, resulting in a final sample of 3839 firms across 30 countries. This sample offers an appropriate context, as the selected countries display variation in both government R&D support and government digital capacity, allowing us to examine how government-level factors influence the relationship between firm underperformance and R&D intensity.
Table 1 reports the distribution of firms by country and industry (two-digit NACE). The sample spans 30 countries. The five countries with the largest number of firms are China, Japan, the United States, Germany, and France. In terms of industry composition, the five most represented manufacturing industries are: Manufacture of computer, electronic and optical products (NACE 26), Manufacture of machinery and equipment n.e.c. (NACE 28), Manufacture of chemicals and chemical products (NACE 20), Manufacture of basic pharmaceutical products and pharmaceutical preparations (NACE 21), and Manufacture of electrical equipment (NACE 27).
4.2 Measurements
4.2.1 Dependent variable
The dependent variable is annual R&D intensity, measured as the ratio (in percentage) of a firm’s R&D expenditures to its operating revenue which is a recognised and established measurement in R&D literature (see Gao et al., 2023; O'Brien & David, 2014).
4.2.2 Independent variables
Firm performance is measured using return on assets (ROA), a widely adopted metric that allows for consistency and comparability across studies. ROA is used in research for examining the relationship between firm performance, financial outcomes, risk-taking, and R&D activity (Desai, 2016; Han, 2023). To construct aspiration levels, we follow the BTOF and adopt a social comparison logic, whereby a firm benchmarks itself against comparable peers (Yu et al., 2019; Zhang et al., 2024). Although aspirations can be based on a firm’s own past performance, we focus on social aspirations because published studies have shown that social aspiration provides the most salient benchmark for what firms are expected to achieve (Yu et al., 2019). Peer performance shapes firm behaviour through competitive and institutional pressures, making it a consequential reference point for managers and external stakeholders (Kim et al., 2015). Additionally, prior works show that when historical and social aspirations are combined, social benchmarks typically dominate, and peer-referenced shortfalls often elicit stronger organisational responses (Bromiley & Harris, 2014; Greve, 2003; Harris & Bromiley, 2007). Thus, in accordance with literature (e.g., Yu et al., 2019; Zhang et al., 2024), we define a firm’s aspiration as the average ROA of firms within the same four-digit NACE industry code, reflecting the importance of industry similarity in forming reference groups (Reger & Huff, 1993; Zhang et al., 2024). Based on these measures, we construct underperformance, calculated as the absolute difference between a firm’s ROA and its aspiration level (Audia & Greve, 2006; Chen & Miller, 2007), with higher values indicating greater performance shortfalls relative to peers. The value of the variable underperformance is equal to zero when a firm’s ROA is equal to or greater than the aspiration level. We also lag underperformance by one year to capture how performance shortfalls influence subsequent R&D intensity.
4.2.3 Moderating variables
We measure government R&D support using the OECD’s R&D Tax Expenditures and Direct Government Funding of BERD database. Among the various indicators provided, we focus specifically on direct government funding of Business Enterprise R&D (BERD) as a percentage of total BERD, as this best captures the extent to which governments provide direct R&D support to firms. This variable is expressed as a percentage, reflecting the proportion of business R&D expenditures financed by public funds. In our sample, the level of government support varies substantially, from a low of 0.46% (Croatia in 2014 and 2015) to a high of 19.57% (Romania in 2012). In our models, we lag government R&D support by one year so that support conditions precede subsequent firm R&D outcomes.
Government digital capacity is measured using the EGDI which evaluates how effectively a country leverages digital technologies to enhance public service delivery and promote social inclusion. The EGDI scores range from 0 to 1 and are grouped into four tiers: very high (0.75–1.00), high (0.50–0.7499), medium (0.25–0.4999), and low (0.00–0.2499). This classification provides a standardised basis for cross-country comparison of digital governance capacity. For example, in our sample, EGDI values range from 0.45872 (South Africa in 2011) to 0.9758 (Denmark in 2020), reflecting variation in digital capacity across countries. In our models, we also lag government digital capacity by one year so that digital conditions precede subsequent firm R&D outcomes.
4.2.4 Control variables
We include several control variables grounded in prior research to account for factors that may influence the results. We control for two broad categories of variables. First, for firm-level characteristics, we control for firm size, measured as the log of the number of employees, as reported in the Orbis database. We include operational efficiency, calculated as cash flow divided by operating revenue, to capture a firm’s internal financing ability and efficiency in turning operations into liquidity (Lee et al., 2019). In addition, we control for shareholder funds, as studies suggest equity levels can both enable and constrain a firm’s R&D investments (Jun et al., 2023). We also control for slack resources, which refer to excess capacity that can be allocated toward innovation. Following Tognazzo et al. (2016) and Yildiz et al. (2023), we measure slack using the current ratio, which reflects a firm’s ability to meet short-term obligations and indicates available liquidity for strategic investments. Finally, we include capital, given its importance in supporting financially intensive activities such as R&D (Manolova et al., 2014).
Second, at the country level, we rely on the World Bank database to obtain data for macro-level control variables. We control for market size, measured as the log of a country’s annual GDP, which captures the scale of economic activity and potential market opportunities. We also control for population size, measured as the log of total population, given population influences firm innovation behaviour and the availability of human capital (Strulik, 2005). We include political stability as a control, as it plays a critical role in shaping the institutional environment, affecting investment risk, regulatory predictability, and transaction costs for firms operating across borders (Goel & Nelson, 2021). Further, we control for trade openness using World Bank data on trade as a percentage of GDP for each country-year. Trade openness is important because exposure to international markets changes firms’ competitive pressure and access to foreign knowledge, which can independently influence innovation and R&D intensity (Queiroz et al., 2025). Additionally, we control for country patent by using data from the WIPO database, measured as the logarithm of patent applications per million residents. Finally, we address unobserved heterogeneity by including firm fixed effects and year fixed effects. In addition, we include two-digit NACE industry-by-year fixed effects to control for time-varying shocks common to firms within the same industry in a given year.
Table 2 provides a summary of variables.
4.3 Analysis
To test our hypotheses, we conducted a panel data analysis using a 10-year dataset spanning the period from 2011 to 2020. Panel data methodology enables the simultaneous examination of both cross-sectional (differences across firms) and longitudinal (differences over time) variation, offering a more comprehensive view of firm behaviour. A key advantage of panel data lies in its ability to account for unobserved heterogeneity, firm-specific or time-invariant characteristics that, if omitted, could bias the estimates. As highlighted by Galindo and Méndez (2014), panel models are suited for isolating the influence of variables that remain constant over time but vary across entities, or vice versa. Moreover, panel data enhances estimation precision by increasing degrees of freedom, reducing multicollinearity, and improving variability in explanatory variables (Baltagi, 2008).
To select the appropriate estimation strategy, we evaluate both fixed effects and random effects models. We apply the Hausman specification test (Hausman, 1978) to assess the consistency of the estimators. The test returned a statistically significant result, so we reject the null hypothesis in favour of the fixed effects model. Additionally, we winsorise the main predictor variables (underperformance, government R&D support, and government digital capacity) at the 1st and 99th percentiles to reduce the influence of extreme observations. We then standardise these variables to enhance comparability across measures and to facilitate interpretation of interaction effects and standardised coefficients.
5 Results
Table 3 reports the descriptive statistics and bivariate correlations for all variables. The Variance Inflation Factor (VIF) analysis shows that all VIF values fall below the conventional threshold of 10 (e.g., Hair et al., 2010).
Table 4 presents the results of the regression analyses with R&D intensity as the dependent variable. Model 1 includes the fixed effects and control variables: firm size, operational efficiency, shareholder funds, slack, capital, market size, population size, political stability, trade openness, and country patent. Model 2 introduces the independent variable, underperformance. Across all models, the F-statistics are statistically significant, suggesting strong overall model fit and satisfactory explanatory power.
In Hypothesis 0 (baseline), we predict that underperformance is positively associated with R&D intensity. This hypothesis is tested in Model 2 of Table 4, where the coefficient for underperformance is positive and statistically significant \(\beta = 0.117, p<0.01\). This result supports Hypothesis 0, indicating that the greater the performance shortfall, the more firms increase R&D intensity. In other words, the magnitude of the gap appears to increase the perceived need for R&D-led recovery, driving firms to intensify their R&D efforts accordingly.
Table 5 reports the results of the moderation analyses. Hypothesis 1 proposed that government R&D support positively moderates the relationship between underperformance and R&D intensity: the positive effect of underperformance on R&D intensity would be stronger in countries with higher levels of government support for R&D. This hypothesis is tested in Model 3 (Table 5), which includes an interaction term between underperformance and government R&D support. The interaction effect is positive and statistically significant (\(\beta =0.059, p<0.01\)), supporting Hypothesis 1. To further understand the size and nature of this interaction, we examine how the effect of underperformance on R&D intensity varies across levels of government R&D support. Following standard recommendations, we probe the interaction using post-estimation simple-slope (conditional marginal effect) tests (Aiken et al., 1991; Dawson, 2014; Lam et al., 2019). The conditional effect of underperformance on R&D intensity is positive but not statistically significant when government R&D support is low \((-1 SD; \beta = 0.0366, p = 0.190, 95\% CI [-0.0181, 0.0913]).\) At the mean level of government R&D support, the conditional effect is positive and statistically significant \((\beta = 0.0961, p < 0.01, 95\% CI [0.0600, 0.1321]\), and it remains positive and statistically significant when government R&D support is high \((+1 SD; \beta = 0.1555,p < 0.01, 95\% CI [0.1157, 0.1953]).\) Importantly, the simple slope at high government support is significantly larger than the slope at low government support \(\Delta \beta = 0.1189, SE = 0.0321, z = 3.70, p < 0.01; 95\% CI [0.0560, 0.1819]).\) These results suggest that while underperformance generally leads to higher R&D investment, the effect is substantially stronger in countries with more robust government support for R&D, highlighting the role of national policy in shaping firm-level strategic responses. The interaction effect is also illustrated in Fig. 1, which plots the relationship between underperformance and R&D intensity at low and high levels of government R&D support. The figure shows that the positive relationship between underperformance and R&D intensity is substantially steeper when government support is high. This suggests that firms experiencing performance shortfalls are significantly more likely to intensify their R&D investments in policy environments where governments actively support R&D.
Interaction effect: Underperformance × government R&D support
Hypothesis 2 predicts that the positive effect of underperformance on R&D intensity would be stronger in countries with higher levels of government digital capacity. This hypothesis is tested in Model 4 (Table 5), which includes an interaction term between underperformance and government digital capacity. The interaction effect is positive and statistically significant \((\beta =0.061, p<.01\)), providing support for Hypothesis 2. We also assess how the effect of underperformance on R&D intensity varies with government digital capacity by estimating conditional (simple) slopes at different levels of digital capacity. Using post-estimation simple-slope tests, we find that underperformance has a positive and statistically significant association with R&D intensity when government digital capacity is low \((-1 SD; \beta = 0.0562, p< 0.05, 95\%\)CI [0.0057, 0.1067]), at its mean \((\beta = 0.1172\), \(p < 0.01, 95\% CI [0.0830, 0.1515])\), and when it is high \((+1 SD; \beta = 0.1783, p < 0.01, 95\% CI [0.1278, 0.2288])\). Moreover, the slope at high digital capacity is significantly larger than the slope at low digital capacity \(\left(\Delta \beta = 0.1221, SE = 0.0379, z = 3.23, p< 0.01; 95\% CI \left[0.0479, 0.1963\right]\right),\) consistent with the interpretation that government digital capacity strengthens the positive relationship between underperformance and firms’ R&D intensity. Figure 2 illustrates this moderation effect, showing that the slope of the relationship steepens under conditions of strong digital capacity, indicating that firms in such environments are more responsive to underperformance through increased R&D investment.
Interaction effect: Underperformance × government digital capacity
Hypothesis 3 predicts that the positive moderating effect of government R&D support on the relationship between underperformance and R&D intensity is contingent on the level of government’s digital capacity. We argue that higher digital capacity strengthens the moderating role of government R&D support, such that the increase in R&D investment in response to underperformance is more pronounced when both government R&D support and digital capacity are high. This three-way interaction is tested in Model 5 (Table 5), which includes the interaction term between underperformance, government R&D support, and government digital capacity. The coefficient for the three-way interaction is positive and statistically significant (\(\beta =0.070, p<0.01\)), supporting Hypothesis 3. However, as established in the interaction literature (e.g., Lam et al., 2019), the sign and significance of the three-way term alone do not convey the full nature of the effect. To clarify the pattern, Fig. 3 plots the interaction, illustrating that the moderating effect of government R&D support on the underperformance–R&D intensity link is conditional on digital capacity. As shown, the association between underperformance and R&D intensity is strongest when both government R&D support and digital capacity are high, indicating that higher digital capacity strengthens the effectiveness of government support in eliciting firms’ R&D responses to underperformance. To further probe the three-way interaction, we computed conditional (simple) slopes of underperformance at the four corner combinations of government R&D support and government digital capacity \(\left(\pm 1 SD\right)\). The conditional slope is largest and statistically significant when both government R&D support and digital capacity are high (+1SD/+1SD; β=0.2268, p < 0.01, 95% CI[0.1721, 0.2814]), whereas the slopes under the mixed conditions (high support/low digital capacity and low support/high digital capacity) are small and not statistically significant. When both moderators are low, the conditional slope remains positive and statistically significant \((-1 SD/-1 SD;\beta = 0.1114, p< 0.05, 95\% CI [0.0082, 0.2145])\). Importantly, a difference-in-differences test confirms that the strengthening effect of government R&D support on the underperformance–R&D intensity relationship is significantly greater when digital capacity is high rather than low \((\Delta = 0.2788, SE = 0.0935, z = 2.98, p< 0.01; 95\% CI [0.0956, 0.4620])\), consistent with the argument that digital capacity strengthens the positive moderating role of government support.
Interaction effect: Underperformance × government R&D support × government digital capacity
The control variables also yield noteworthy insights. Firm size and operational efficiency consistently exhibit negative and statistically significant effects on R&D intensity, suggesting that larger or more operationally efficient firms may be less inclined to invest in R&D, possibly due to established routines or risk aversion. In contrast, slack resources exhibit a positive and statistically significant association with R&D intensity, indicating that the availability of excess resources can facilitate or encourage innovation investment. Among the country-level controls, market size shows a positive and statistically significant relationship with R&D intensity, while population size is negative and statistically significant. Trade openness is also negative and statistically significant. By contrast, political stability and country patent are not statistically significant, and shareholder funds and capital are also not statistically significant. These findings underscore the influence of the external environment on firms’ strategic decisions regarding R&D and innovation.
5.1 Robustness tests
To test the robustness of our results, we conducted several additional tests. First, we re-estimate our models using correlated random effects (CRE) (Schunck, 2013; Wooldridge, 2019). CRE provides an alternative to fixed effects that allows for unobserved unit heterogeneity while relaxing the strict assumption that this heterogeneity is unrelated to the regressors. In our application, CRE is useful because it offers a complementary way to account for persistent firm-level differences and to assess whether our results depend on the fixed effects estimation. Results from the CRE estimates are consistent with the baseline fixed effects results, and our main findings, including the interaction patterns, remain unchanged.
Second, we re-estimate the models using Driscoll-Kraay standard errors (Driscoll & Kraay, 1998), which are robust to heteroskedasticity, autocorrelation, and cross-sectional dependence. Under this more conservative specification, the main effect of underperformance, the moderating effect of government digital capacity, and the three-way interaction remain statistically significant and consistent in sign with the baseline estimates. Only the interaction between underperformance and government R&D support is not statistically significant, although its sign remains unchanged. This attenuation is not unexpected, as Driscoll-Kraay standard errors tend to be more conservative, and interaction terms can be particularly sensitive when collinearity inflates standard errors (Hoechle, 2007). Overall, with the significant three-way interaction, these results reinforce our argument that the moderating role of government R&D support is conditional on government digital capacity.
Third, we assess whether our findings are sensitive to alternative operationalisations of performance aspiration levels. Prior BTOF research has used different approaches to capture firms’ reference points, including social aspirations measured as the median performance of industry peers to reduce sensitivity to extreme values (e.g., Zhang et al., 2024), as well as hybrid measures that combine a firm’s own past performance with the performance of industry peers (e.g., Gao et al., 2023). Following these alternatives, we re-estimated all models using alternative aspiration measures and reconstructed the underperformance variable accordingly, while holding all other model specifications unchanged. First, we operationalised social aspiration using the median ROA of firms within the same four-digit NACE industry code and year, recalculated underperformance based on this median-based aspiration, and repeated the full analyses; the results remained consistent with our main predictions. Second, we constructed an overall aspiration level as a weighted average of historical and social aspirations (where historical aspiration is the firm’s own ROA in the prior year and social aspiration is the average ROA of firms in the same four-digit NACE industry-year) using weights established in prior work; we created one measure with 20% historical and 80% social aspiration (O’Brien and David, 2014) and another with 30% historical and 70% social aspiration (Zhang et al., 2023). Using each weighted aspiration level, we reconstructed underperformance and re-ran all regressions, with all the results consistent with the prior findings. Third, because our multi-country setting raises the possibility that firms may rely more heavily on geographically proximate reference groups, we defined an alternative aspiration level based on the average ROA of firms operating in the same four-digit NACE industry, within the same continent and year, reconstructed underperformance using this continent-bounded benchmark, and re-estimated all models; results again remained consistent with the main analyses, indicating that our findings are robust to different measurements of aspiration levels.
Finally, we extended the dataset to include 2021–2023 and re-estimated all models. This sample, covering 2011–2023, includes both the pandemic period and the early post-pandemic recovery. The results remain consistent in sign, magnitude, and significance indicating that conclusions are robust to the inclusion of COVID-related economic shocks. Taken together, these robustness tests strengthen confidence in our findings.
6 Discussion
This study examines how the national institutional environment conditions firms’ R&D response to underperformance. Building on the BTOF, we argue and empirically find that an increase in R&D intensity as a response to underperformance is not only a function of aspiration–performance gaps and firm-internal contingencies, but also of country-level policy and information infrastructures. In particular, government R&D support and government digital capacity each strengthen the positive association between underperformance and R&D intensity, and they also do so in combination. In doing so, we reposition the question from whether underperformance triggers R&D to how strongly firms scale R&D when performance falls short under different institutional conditions. With this study, we make several theoretical and practical contributions.
6.1 Theoretical contributions
Consistent with foundational BTOF tenets, we reaffirm that underperformance elicits problemistic search, increasing firms’ R&D intensity (e.g., Bromiley & Washburn, 2011; Chen et al., 2025; Zhao et al., 2025). We advance the BTOF by explicitly embedding this mechanism in the institutional environment in which firms operate. In particular, we develop a cross-level framework in which firm-level performance feedback unfolds within national regimes that shape access to resources and information. We show that direct government R&D support expands the feasible set of resource commitments and confers legitimacy on underperforming firms, enabling them to translate intent into funded R&D. This offers a behavioural contingency perspective on the “crowd-in vs. crowd-out” policy debate and links to behavioural additionality (Cunningham et al., 2016). We also theorise and find that government digital capacity (distinct from generic bureaucratic efficiency) enhances the timeliness, transparency, and scalability of search, thereby strengthening the translation of underperformance into strategic R&D moves. Together, these results extend the BTOF beyond firm-internal moderators (e.g., slack) to external, country-level boundary conditions that systematically shape the intensity of search.
Second, we advance a configurational argument about how government-level factors shape problemistic search, where government R&D support and government digital capacity operate as institutional complements. The core theoretical implication is that the effect of government support is implementation-contingent. Government R&D support can expand firms’ feasible resource set and relax constraints, but the extent to which firms respond behaviourally depends on whether the government has complementary capabilities that make policy reliable, recognisable, and straightforward to mobilise. In this view, government digital capacity is an institutional capability that conditions whether firms can translate government programmes into realised support under bounded rationality and constrained attention, which are central assumptions in the BTOF (Cyert & March, 1963). This complementarity perspective reframes cross-national heterogeneity in underperformance-induced R&D as variation in the configuration of institutional enablers. Similar levels of government R&D support can yield divergent firm responses because government digital capacity changes the perceived accessibility, predictability, and strategic usefulness of that support. More broadly, by foregrounding implementation contingency alongside the level of government R&D support, our argument clarifies when government R&D support strengthens problemistic search and, in turn, R&D intensity, thereby helping reconcile mixed prior evidence by identifying government digital capacity as the key conditioning factor (Martínez-Noya & Valdés-Llaneza, 2025).
In summary, by integrating government levers into performance-feedback models, we recast problemistic search as a cross-level outcome co-produced by managerial responses and institutional structures that shape firms’ option sets and frictions. Conceptually, we enrich the BTOF with testable, operationalisable constructs for government support and digital capacity; empirically, our multi-country panel underscores that these mechanisms travel beyond single-country settings. Overall, we advance an institutionally embedded view of adaptive search: how strongly firms convert underperformance into R&D depends on the configuration of support and information infrastructures (digital capacity) in the environments in which they are embedded.
6.2 Practical implications
This study also provides several practical implications. Since underperforming firms are especially responsive to financial stimuli, support schemes should explicitly lower entry barriers for distressed firms (Qu et al., 2017; Xu et al., 2021). Rather than treating funding as a stand-alone instrument, governments should design integrated R&D policies that bundle financial resources with high-capacity digital delivery systems (Liu et al., 2024). When the scale of support is matched to firms’ performance needs and is delivered through low-friction, data-rich channels, public money travels further, raising R&D intensity via stronger behavioural additionality.
The results also highlight that government digital capacity increases the effectiveness of R&D support. In governments with high digital maturity, e-government platforms make R&D support more visible and easier to mobilise by streamlining eligibility checks and applications, increasing transparency about rules and timelines, and lowering administrative burdens, which in turn enables firms to access funding more quickly and respond with more timely and targeted R&D investments (Bosio et al., 2023; Zhang & Kaur, 2024). Concretely, real-time eligibility screening, searchable grant catalogues with APIs, and automated disbursement/compliance tracking lower search and transaction costs, enabling faster and more strategic R&D deployment. Investing in digital platforms is therefore both an administrative upgrade and a policy mechanism that multiplies the R&D impact of financial support. For governments, this means that R&D policy must treat financial and digital infrastructures as complementary levers. Budgeting and evaluation should be co-developed, allocating funds to programs in tandem with the digital upgrades that make them usable and auditable. Particularly in low-digital-capacity environments, even generous support programs may underperform if access remains opaque or bureaucratic. By contrast, in environments with advanced digital systems, even modest funding can be converted quickly into projects with clear milestones and learning loops. Thus, strategic alignment of funding mechanisms with digital capability becomes particularly important to ensure that resources are not only disbursed but also absorbed productively by the intended firms.
6.3 Limitations and future research directions
While our findings are robust and offer insightful implications, our study is not without limitations. First, our measures for government R&D support and digital capacity are at the country level. Although this is consistent with our cross-level theorizing, it abstracts from within-country heterogeneity in program design, administrative practices, and firms’ ability to engage with government support. Future studies could gain precision by using firm-level data on the actual receipt, amount, and type of support (e.g., grants versus other instruments) and by developing firm-level measures of perceived digital government effectiveness, which may vary even within the same country due to differences in firm capabilities, location, or prior experience interacting with government programs (Dimos & Pugh, 2016; Hottenrott & Lopes-Bento, 2020; Malodia et al., 2021). Future research could also differentiate forms of government support by separating instruments that primarily relax financing constraints from those that explicitly encourage collaboration and learning, which aligns with behavioural additionality arguments (Cunningham et al., 2016).
Second, our dependent variable is R&D intensity, an input measure of innovation (Hagedoorn & Cloodt, 2003). While appropriate for capturing the allocation of resources to search, it cannot speak directly to whether such investments translate into innovation outputs or broader performance improvements. Future research could extend our model to examine outputs such as patent filings, citation-weighted patents, new product introductions, or productivity outcomes. This would clarify whether supportive institutional contexts primarily facilitate spending increases, improve the efficiency of converting inputs into outputs, or both. It would also help connect performance feedback dynamics more explicitly to downstream innovation performance.
Third, our cross-national setting includes 30 countries, which provides meaningful variation but also limits generalisability. Our results should be interpreted as evidence from this specific set of institutional environments and time rather than as universal effects. Future studies could expand coverage to a larger number of countries, including emerging and lower-income contexts where institutional voids and administrative capacity constraints may be more severe, potentially altering the strength or form of the moderation effects (North, 1990; Rothstein & Teorell, 2008).
Fourth, our sample is focused on manufacturing firms. Given the different nature of innovation in services and digital-intensive sectors, future research should examine whether the same institutional levers condition performance feedback responses where innovation is less R&D-based and more driven by data, platform ecosystems, or process redesign. Extending the analysis to non-manufacturing sectors would strengthen external validity and refine the boundary conditions of our arguments.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and interpreting interactions. Sage.
Akpan-Obong, P. I., Trinh, M. P., Ayo, C. K., & Oni, A. (2023). E-governance as good governance? Evidence from 15 West African countries. Information Technology for Development, 29(2–3), 256–275.
Argote, L., & Greve, H. R. (2007). A behavioral theory of the firm—40 years and counting: Introduction and impact. Organization Science, 18(3), 337–349.
Audia, P. G., & Greve, H. R. (2006). Less likely to fail: Low performance, firm size, and factory expansion in the shipbuilding industry. Management Science, 52(1), 83–94.
Bacharach, S. B. (1989). Organizational theories: Some criteria for evaluation. Academy of Management Review, 14(4), 496–515.
Bae, S. C., Park, B. J., & Wang, X. (2008). Multinationality, R&D intensity, and firm performance: Evidence from US manufacturing firms. Multinational Business Review, 16(1), 53–78.
Baltagi, B. H. (2008). Forecasting with panel data. Journal of Forecasting, 27(2), 153–173.
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120.
Baum, J. A., & Dahlin, K. B. (2007). Aspiration performance and railroads’ patterns of learning from train wrecks and crashes. Organization Science, 18(3), 368–385.
Baum, J. A., Rowley, T. J., Shipilov, A. V., & Chuang, Y. T. (2005). Dancing with strangers: Aspiration performance and the search for underwriting syndicate partners. Administrative Science Quarterly, 50(4), 536–575.
Bhaumik, S., Driffield, N., Lavoratori, K., Song, M., & Vahter, P. (2025). Multinational strategy, institutions and spillovers: The role of institutions in knowledge spillovers in emerging markets. The Journal of Technology Transfer. https://doi.org/10.1007/s10961-025-10230-w
Bosio, E., Hayman, G., & Dubosse, N. (2023). The investment case for e-government procurement: A cost–benefit analysis. Journal of Benefit-Cost Analysis, 14(S1), 81–107.
Bromiley, P., & Harris, J. D. (2014). A comparison of alternative measures of organizational aspirations. Strategic Management Journal, 35(3), 338–357.
Bromiley, P., & Washburn, M. (2011). Cost reduction vs innovative search in R&D. Journal of Strategy and Management, 4(3), 196–214.
Buisseret, T. J., Cameron, H., & Georghiou, L. (1995). What difference does it make? Additionality in the public support of R&D in large firms. International Journal of Technology Management, 10(4–6), 587–600.
Busse, C., Kach, A. P., & Wagner, S. M. (2017). Boundary conditions: What they are, how to explore them, why we need them, and when to consider them. Organizational Research Methods, 20(4), 574–609.
Carnes, C. M., Hitt, M. A., Sirmon, D. G., Chirico, F., & Huh, D. W. (2022). Leveraging resources for innovation: The role of synchronization. Journal of Product Innovation Management, 39(2), 160–176.
Chen, G., Liang, T., & Chen, X. (2025). How do firms’ performance shortfalls affect digital transformation? The role of peer experience and persistent underperformance. Review of Managerial Science, 19(3), 757–795.
Chen, S., Wang, X., Gan, T., & Gui, G. (2024). The effect of digital government on corporate total factor productivity. PLoS ONE, 19(9), e0308093.
Chen, W. R. (2008). Determinants of firms’ backward-and forward-looking R&D search behavior. Organization Science, 19(4), 609–622.
Chen, W. R., & Miller, K. D. (2007). Situational and institutional determinants of firms’ R&D search intensity. Strategic Management Journal, 28(4), 369–381.
Clarysse, B., Wright, M., & Mustar, P. (2009). Behavioural additionality of R&D subsidies: A learning perspective. Research Policy, 38(10), 1517–1533.
Cunningham, P., Gök, A., & Laredo, P. (2016). The impact of direct support to R&D and innovation in firms. In Handbook of innovation policy impact (pp. 54–107). Edward Elgar Publishing.
Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. Prentice-Hall.
Dai, L., Eden, L., & Beamish, P. W. (2023). The timing and mode of foreign exit from conflict zones: A behavioral perspective. Journal of International Business Studies, 54(6), 1090–1104.
David, P. A., Hall, B. H., & Toole, A. A. (2000). Is public R&D a complement or substitute for private R&D? A review of the econometric evidence. Research Policy, 29(4–5), 497–529.
Dawson, J. F. (2014). Moderation in management research: What, why, when, and how. Journal of Business and Psychology, 29(1), 1–19.
Desai, V. M. (2016). The behavioral theory of the (governed) firm: Corporate board influences on organizations’ responses to performance shortfalls. Academy of Management Journal, 59(3), 860–879.
Dhaoui, I. (2022). E-government for sustainable development: Evidence from MENA countries. Journal of the Knowledge Economy, 13(3), 2070–2099.
Dimos, C., & Pugh, G. (2016). The effectiveness of R&D subsidies: A meta-regression analysis of the evaluation literature. Research Policy, 45(4), 797–815.
Doh, S., & Kim, B. (2014). Government support for SME innovations in the regional industries: The case of government financial support program in South Korea. Research Policy, 43(9), 1557–1569.
Dong, J. Q. (2021). Technological choices under uncertainty: Does organizational aspiration matter? Strategic Management Journal, 42(5), 898–916.
Driscoll, J. C., & Kraay, A. C. (1998). Consistent covariance matrix estimation with spatially dependent panel data. Review of Economics and Statistics, 80(4), 549–560.
Edwards, J. R., & Berry, J. W. (2010). The presence of something or the absence of nothing: Increasing theoretical precision in management research. Organizational Research Methods, 13(4), 668–689.
Ehls, D., Polier, S., & Herstatt, C. (2020). Reviewing the field of external knowledge search for innovation: Theoretical underpinnings and future (re-) search directions. Journal of Product Innovation Management, 37(5), 405–430.
Falk, M. (2012). Quantile estimates of the impact of R&D intensity on firm performance. Small Business Economics, 39(1), 19–37.
Galindo, M. Á., & Méndez, M. T. (2014). Entrepreneurship, economic growth, and innovation: Are feedback effects at work? Journal of Business Research, 67(5), 825–829.
Gao, R., Lu, J. W., Hu, H. W., & Martin, G. (2023). A tale of two distractions: How institutional forces influence R&D-based problemistic search in transition economies. Journal of Product Innovation Management, 40(5), 657–678.
Gavetti, G., Greve, H. R., Levinthal, D. A., & Ocasio, W. (2012). The behavioral theory of the firm: Assessment and prospects. The Academy of Management Annals, 6(1), 1–40.
Goel, R. K., & Nelson, M. A. (2021). How do firms use innovations to hedge against economic and political uncertainty? Evidence from a large sample of nations. The Journal of Technology Transfer, 46(2), 407–430.
Goyal, L. (2025). Institutional history, negative performance feedback, and R&D search: A nexus of the imprinting and behavioral perspectives. Academy of Management Journal.
Greve, H. R. (1998). Performance, aspirations, and risky organizational change. Administrative Science Quarterly, 43(1), 58–86.
Greve, H. R. (2003). A behavioral theory of R&D expenditures and innovations: Evidence from shipbuilding. Academy of Management Journal, 46(6), 685–702.
Greve, H. R., & Gaba, V. (2017). Performance feedback in organizations and groups: Common themes. In L. Argote & J. M. Levine (Eds.), The Oxford handbook of group and organizational learning (pp. 315–335). Oxford University Press.
Guan, J., & Yam, R. C. (2015). Effects of government financial incentives on firms’ innovation performance in China: Evidences from Beijing in the 1990s. Research Policy, 44(1), 273–282.
Guellec, D. (2003). The impact of public R&D expenditure on business R&D. Economics of Innovation and New Technology, 12(3), 225–243.
Guo, D., Guo, Y., & Jiang, K. (2016). Government-subsidized R&D and firm innovation: Evidence from China. Research Policy, 45(6), 1129–1144.
Guo, Y., Chen, J., & Liu, Z. (2022). Government responsiveness and public acceptance of big-data technology in urban governance: Evidence from China during the COVID-19 pandemic. Cities, 122, 103536.
Hagedoorn, J., & Cloodt, M. (2003). Measuring innovative performance: Is there an advantage in using multiple indicators? Research Policy, 32(8), 1365–1379.
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th ed.). Prentice-Hall.
Han, S. (2023). The effect of performance feedback on strategic alliance formation and R&D intensity. European Management Journal, 41(5), 709–719.
Harris, J., & Bromiley, P. (2007). Incentives to cheat: The influence of executive compensation and firm performance on financial misrepresentation. Organization Science, 18(3), 350–367.
Hausman, J. A. (1978). Specification tests in econometrics. Econometrica, 46(6), 1251–1271.
Hillman, A., & Keim, G. (1995). International variation in the business-government interface: Institutional and organizational considerations. Academy of Management Review, 20(1), 193–214.
Hoechle, D. (2007). Robust standard errors for panel regressions with cross-sectional dependence. The Stata Journal, 7(3), 281–312.
Hottenrott, H., & Lopes-Bento, C. (2020). R&D subsidies and firm-level innovation performance. DICE Report, 18(4), 20–25.
Hu, S., Gu, Q., & Xia, J. (2022). Problemistic search of the embedded firm: The joint effects of performance feedback and network positions on venture capital firms’ risk taking. Organization Science, 33(5), 1889–1908.
Huang, L., He, L., & Yang, G. (2021). Performance shortfalls and R&D investment change: Aspirations, actions, and expectations. Sustainability, 13(6), 3006.
Iyer, D. N., & Miller, K. D. (2008). Performance feedback, slack, and the timing of acquisitions. Academy of Management Journal, 51(4), 808–822.
Jia, L., Nam, E., & Chun, D. (2021). Impact of Chinese government subsidies on enterprise innovation: Based on a three-dimensional perspective. Sustainability, 13(3), 1288.
Jun, C., Qiyuan, L., Xiaofang, M., & Zhang, F. F. (2023). Board gender diversity and cost of equity: Evidence from mandatory female board representation. International Review of Economics & Finance, 88, 501–515.
Kang, K. N., & Park, H. (2012). Influence of government R&D support and inter-firm collaborations on innovation in Korean biotechnology SMEs. Technovation, 32(1), 68–78.
Kim, J. Y., Finkelstein, S., & Haleblian, J. (2015). All aspirations are not created equal: The differential effects of historical and social aspirations on acquisition behavior. Academy of Management Journal, 58(5), 1361–1388.
Klein, K. J., Dansereau, F., & Hall, R. J. (1994). Levels issues in theory development, data collection, and analysis. Academy of Management Review, 19(2), 195–229.
Kotiloglu, S., Chen, Y., & Lechler, T. (2021). Organizational responses to performance feedback: A meta-analytic review. Strategic Organization, 19(2), 285–311.
Kuusela, P., Keil, T., & Maula, M. (2017). Driven by aspirations, but in what direction? Performance shortfalls, slack resources, and resource-consuming vs. resource-freeing organizational change. Strategic Management Journal, 38(5), 1101–1120.
Lam, L. W., Chuang, A., Wong, C. S., & Zhu, J. N. (2019). A typology of three-way interaction models: Applications and suggestions for Asian management research. Asia Pacific Journal of Management, 36(1), 1–16.
Lantz, J. S., & Sahut, J. M. (2005). R&D investment and the financial performance of technological firms. International Journal of Business, 10(3), 251.
Lee, J., Kwon, H. B., & Pati, N. (2019). Exploring the relative impact of R&D and operational efficiency on performance: A sequential regression-neural network approach. Expert Systems with Applications, 137, 420–431.
Lenihan, H., Mulligan, K., Doran, J., Rammer, C., & Ipinnaiye, O. (2024). R&D grants and R&D tax credits to foreign-owned subsidiaries: Does supporting multinational enterprises’ R&D pay off in terms of firm performance improvements for the host economy? The Journal of Technology Transfer, 49(2), 740–781.
Levinthal, D., & March, J. G. (1981). A model of adaptive organizational search. Journal of Economic Behavior & Organization, 2(4), 307–333.
Li, H., & Xu, J. (2024). Impact of digital government on digital transformation of enterprises from the perspective of urban economic sustainable development. Sustainability, 16(7), 2667.
Lim, E. N., & McCann, B. T. (2014). Performance feedback and firm risk taking: The moderating effects of CEO and outside director stock options. Organization Science, 25(1), 262–282.
Liu, P., Zhu, B., Yang, M., & De Baets, B. (2024). Contribution of digital governments to digital transformation of firms: Evidence from China. Technology Analysis & Strategic Management, 1–15.
Madadian, O., & Van den Broeke, M. (2023). R&D investments in response to performance feedback: Moderating effects of firm risk profile and business strategy. Applied Economics, 55(7), 802–822.
Malodia, S., Dhir, A., Mishra, M., & Bhatti, Z. A. (2021). Future of e-Government: An integrated conceptual framework. Technological Forecasting and Social Change, 173, 121102.
Manolova, T. S., Manev, I. M., & Gyoshev, B. S. (2014). Friends with money? Owner’s financial network and new venture internationalization in a transition economy. International Small Business Journal, 32(8), 944–966.
Martínez-Noya, A., & Valdés-Llaneza, A. (2025). Performance feedback and international business research: A review and future directions. International Business Review. https://doi.org/10.1016/j.ibusrev.2025.102426
Micol, F., Battaglia, D., & Ughetto, E. (2025). Private entrepreneurial support organizations in European fintech entrepreneurial ecosystems. The Journal of Technology Transfer, 50(3), 1170–1198.
Nave, E., Torres, P., Querido, A. R., Ferreira, J. J., & Fernandes, G. (2024). Entrepreneurial ecosystems governance: Critical perspectives and steps forward. The Journal of Technology Transfer. https://doi.org/10.1007/s10961-024-10172-9
North, D. C. (1990). Institutions, institutional change and economic performance. Cambridge University Press.
O’Brien, J. P., & David, P. (2014). Reciprocity and R&D search: Applying the behavioral theory of the firm to a communitarian context. Strategic Management Journal, 35(4), 550–565.
OECD. (2020). The OECD digital government policy framework: Six dimensions of a digital government (OECD Public Governance Policy Papers, No. 2).1 OECD Publishing.
Ozturk-Kose, E. (2025). Home country institutions and exports of emerging economy small firms: The role of internal R&D and acquisition of external knowledge. The Journal of Technology Transfer, 50(2), 752–777.
Pearce, J. L., Xin, K. M., Xu, Q. J., & Rao, A. N. (2011). Managers’ context: How government capability affects managers. British Journal of Management, 22(3), 500–516.
Peng, M. W. (2003). Institutional transitions and strategic choices. Academy of Management Review, 28(2), 275–296.
Petrin, T. (2018). A literature review on the impact and effectiveness of government support for R&D and innovation. ISIGrowth.
Posen, H. E., Keil, T., Kim, S., & Meissner, F. D. (2018). Renewing research on problemistic search—A review and research agenda. The Academy of Management Annals, 12(1), 208–251.
Qu, J. A., Cao, J., Wang, X. T., Tang, J. X., & Bukenya, J. O. (2017). Political connections, government subsidies and technical innovation of wind energy companies in China. Sustainability, 9(10), 1812.
Queiroz, S., Vonortas, N., & Canuto, O. (2025). Global changes and effectiveness of innovation policy in Brazil. The Journal of Technology Transfer, 1–20.
Ref, O., & Shapira, Z. U. R. (2017). Entering new markets: The effect of performance feedback near aspiration and well below and above it. Strategic Management Journal, 38(7), 1416–1434.
Reger, R. K., & Huff, A. S. (1993). Strategic groups: A cognitive perspective. Strategic Management Journal, 14(2), 103–123.
Rhee, L., Ocasio, W., & Kim, T. H. (2019). Performance feedback in hierarchical business groups: The cross-level effects of cognitive accessibility on R&D search behavior. Organization Science, 30(1), 51–69.
Rothstein, B. O., & Teorell, J. A. (2008). What is quality of government? A theory of impartial government institutions. Governance, 21(2), 165–190.
Rudy, B. C., & Johnson, A. F. (2016). Performance, aspirations, and market versus nonmarket investment. Journal of Management, 42(4), 936–959.
Schunck, R. (2013). Within and between estimates in random-effects models: Advantages and drawbacks of correlated random effects and hybrid models. The Stata Journal: Promoting Communications on Statistics and Stata, 13(1), 65–76.
Scoresby, R. B., Withers, M. C., & Ireland, R. D. (2021). The effect of CEO regulatory focus on changes to investments in R&D. Journal of Product Innovation Management, 38(4), 401–420.
Seiam, D. A., & Salman, D. (2024). Examining the global influence of e-governance on corruption: A panel data analysis. Future Business Journal, 10(1), 29.
Sengul, M., & Obloj, T. (2017). Better safe than sorry: Subsidiary performance feedback and internal governance in multiunit firms. Journal of Management, 43(8), 2526–2554.
Stephan, U., Uhlaner, L. M., & Stride, C. (2015). Institutions and social entrepreneurship: The role of institutional voids, institutional support, and institutional configurations. Journal of International Business Studies, 46(3), 308–331.
Strulik, H. (2005). The role of human capital and population growth in R&D-based models of economic growth. Review of International Economics, 13(1), 129–145.
Tognazzo, A., Gubitta, P., & Favaron, S. D. (2016). Does slack always affect resilience? A study of quasi-medium-sized Italian firms. Entrepreneurship & Regional Development, 28(9–10), 768–790.
Von Nitzsch, J., Bird, M., & Saiedi, E. (2024). The strategic role of owners in firm growth: Contextualizing ownership competence in private firms. Strategic Entrepreneurship Journal, 18(3), 553–581.
Watanabe, C., Takayama, M., Nagamatsu, A., Tagami, T., & Griffy-Brown, C. (2002). Technology spillover as a complement for high-level R&D intensity in the pharmaceutical industry. Technovation, 22(4), 245–258.
Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of Management Review, 14(4), 490–495.
Williams, C. C. (2023). Formalization of the informal economy: An e-government approach. Springer Nature.
Wooldridge, J. M. (2019). Correlated random effects models with unbalanced panels. Journal of Econometrics, 211(1), 137–150.
Xu, C., & Jin, L. (2024). Effects of government digitalization on firm investment efficiency: Evidence from China. International Review of Economics & Finance, 92, 819–834.
Xu, D., Zhou, K. Z., & Du, F. (2019). Deviant versus aspirational risk taking: The effects of performance feedback on bribery expenditure and R&D intensity. Academy of Management Journal, 62(4), 1226–1251.
Xu, J., Wang, X., & Liu, F. (2021). Government subsidies, R&D investment and innovation performance: Analysis from pharmaceutical sector in China. Technology Analysis & Strategic Management, 33(5), 535–553.
Xue, S., Chen, H., & Wu, J. (2023). Performance feedback and firms’ strategic actions: A cross-national meta-analysis. Management Decision, 61(7), 2201–2230.
Yildiz, H. E., Morgulis‐Yakushev, S., Holm, U., & Eriksson, M. (2023). Directionality matters: Board interlocks and firm internationalization. Global Strategy Journal, 13(1), 90–110.
Yu, W., Minniti, M., & Nason, R. (2019). Underperformance duration and innovative search: Evidence from the high-tech manufacturing industry. Strategic Management Journal, 40(5), 836–861.
Zhang, M., & Kaur, M. (2024). Toward a theory of e-government: Challenges and opportunities, a literature review. Journal of Infrastructure, Policy and Development. https://doi.org/10.24294/jipd.v8i10.7707
Zhang, M., Shui, X., Smart, P., Wang, X., & Chen, J. (2023). Environmental performance feedback and timing of reshoring: Perspectives from the behavioural theory of the firm. British Journal of Management, 34(3), 1238–1258.
Zhang, Z., Gu, X., & Yang, X. (2024). Performance feedback and innovative search strategies: An integrative perspective of motivation and capacity for risk taking. British Journal of Management, 35(4), 1867–1885.
Zhao, J., Cheng, X., & Guo, J. (2025). How do aspirations work? Financial performance shortfalls and firms' green innovation. Business Strategy and the Environment.
Zhou, X., Wang, Z., & Wang, J. (2024). Government support, director characteristics and innovation performance: Evidence from China’s digital creative industry. International Journal of Technology Management, 94(2), 271–290.
Funding
No funding was received for conducting this study.
Author information
Authors and Affiliations
Contributions
All authors contributed at different stages of the study and approved the final manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors have no competing interests to declare that are relevant to the content of this article.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Hamrabadi, A., Khorana, S. & Calik, A. Beyond the firm: how government R&D support and digital capacity shape firms’ R&D response to underperformance. J Technol Transf (2026). https://doi.org/10.1007/s10961-026-10332-z
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1007/s10961-026-10332-z