Unlocking the export potential of small entrepreneurship in Uzbekistan through efficient market infrastructure: a political economy opinion

Uzbekistan's reform decade has produced a revealing paradox. By late 2025 the country counted roughly 1.21 million active small business entities generating about half of gross domestic product, yet small business accounted for only about 32 percent of merchandise exports in the first nine months of 2025, against some 54 percent of imports (National Statistics Committee, 2025). The country is, in effect, a nation of small entrepreneurs who sell briskly at home but seldom abroad.Trade theory predicts such selection: because exporting carries large sunk and fixed costs, only more productive firms cross the border (Roberts and Tybout, 1997;Melitz, 2003;Wagner, 2007). Is that threshold fixed by nature, or can design lower it? Three literatures speak to this question but rarely to one another: work on SME internationalization explains export participation through firm-level attributes; tradefacilitation research estimates how procedural costs shape aggregate flows; political economy explains why efficient policies go unadopted. The gap we address is the absence of an integrated account treating the height of the export selection threshold as an endogenous political outcome rather than a technical parameter or firm characteristic. Our contribution is to specify that linkage as an explicit causal chain, to argue, as a proposition open to test, which tier is likely to be binding, and to derive reform levers from the mechanisms themselves rather than from a generic infrastructure agenda.Uzbekistan is an unusually informative case. Doubly landlocked, its hard-infrastructure penalties cannot be assumed away. It is mid-transition, with a large state-owned sector alongside rapid private entry, rendering distributional conflict visible rather than latent. It is in active WTO accession, a rare chance to observe external commitment as a domestic disciplining device. And it has adopted a complete trade-facilitation statutory framework while firm-level friction persists, separating the implementation gap from the legislative.Across every major domestic indicator small business hovers near half of all activity: 51.5 percent of GDP, 55 percent of market services, 51.8 percent of freight transport (National Statistics Committee, 2025). At the border its footprint collapses to roughly a third. That small firms buy readily from international markets but rarely sell into them points to obstacles on the supply and market-access side, not to any aversion to cross-border trade. This matters because what a country exports shapes what it later learns to export, so a narrow commodity basket forecloses future capability (Hausmann et al., 2007;World Bank, 2025;ADB, 2025).Why, then, do domestically successful firms so rarely export? Certification, foreign standards, customs, trade finance and buyer search are paid before the first shipment leaves. Large firms spread these fixed costs over volume; small firms cannot (Bernard and Jensen, 1999). There is also a discovery cost: learning what one can profitably sell abroad is a quasi-public good, since pioneers are quickly imitated (Hausmann and Rodrik, 2003).We therefore advance the following proposition, to be defended rather than assumed: the decisive variable is the shared infrastructure lowering the fixed cost of a first export. Following Portugal-Perez and Wilson (2012), we distinguish two layers. Hard infrastructure denotes the physical assets of trade: transport corridors, multimodal terminals, cold chain, bonded warehousing. Soft infrastructure denotes the institutional and procedural apparatus: customs and single-window systems, quality and certification infrastructure, trade finance, market intelligence. The hard layer carries a landlocked penalty falling hardest on low-margin small consignments (Arvis et al., 2010); the soft layer is where small firms most lack scale, legal capacity and information. Uzbekistan posts one of the region's highest trade-facilitation implementation rates, yet firms still find trading slow and costly (OECD, 2023;UNECE, 2025).Reform has reached the statute book faster than the warehouse and the border post.That the soft layer should be prioritised for small enterprises follows from cost structure, and we present it as a proposition grounded in that structure rather than as a settled result. Softinfrastructure costs are incurred per firm and per consignment rather than per unit of value: an accreditation, a conformity assessment, a customs declaration and a buyer search cost the same whether a shipment is worth ten thousand dollars or ten million. They are therefore steeply regressive in firm size, whereas hard-infrastructure costs scale more nearly with volume. Delay compounds the asymmetry, since time in transit acts as a substantial ad valorem tariff, bearing hardest on the perishable, low-margin consignments typical of small exporters (Djankov et al., 2010;Hummels and Schaur, 2013). Small firms are also least able to substitute private capacity for missing public provision, lacking in-house legal, compliance and market-intelligence functions (Beck and Demirgüç-Kunt, 2006). Consistent with this, estimated returns to soft infrastructure exceed those to hard infrastructure in lower-income settings (Portugal-Perez and Wilson, 2012), and firm-level evidence across developing countries does not bear out the concern that facilitation gains are captured mainly by large firms (Hoekman and Shepherd, 2015). Recent cross-country firm-level evidence points in the same direction: burdensome customs requirements deter firm participation in international production networks, while regulatory quality, access to credit and adoption of digital communication support it, with favourable effects extending to less productive firms (Reddy and Sasidharan, 2025).We should be explicit about what this argument does not claim. Market infrastructure is not the only constraint on SME exports, nor the largest in every sector. Firm productivity and managerial capability remain first-order determinants of export entry, and much of the observed selection would persist under any infrastructure regime (Melitz, 2003;Wagner, 2007). Innovation capacity and product quality determine whether a firm has anything a foreign buyer wants; access to finance constrains working capital independently of trade finance institutions (Beck and Demirgüç-Kunt, 2006); human capital shapes absorptive capacity for foreign standards; digital infrastructure and macroeconomic conditions shape entrepreneurial activity more broadly (Abid, 2025); and exchange-rate competitiveness affects export profitability at the margin.Our claim is narrower and remains a proposition. These explanations are complementary, and several are partly endogenous to infrastructure. What distinguishes the infrastructure account is that it identifies a constraint the state can relax quickly, whereas productivity and human capital respond only over long horizons. Their relative weight is an open question this Opinion does not settle.Figure 1 organises these observations into a single analytical sequence. It is a diagnostic device, not an estimated model, built by conceptual synthesis following an established procedure: identifying the causal questions the phenomenon raises, mapping the traditions answering each, and ordering them by direction of determination (Jabareen, 2009;Snyder, 2019). Four questions generate the four tiers. What screens firms out? A threshold set by the fixed and discovery costs of exporting (Melitz, 2003;Hausmann and Rodrik, 2003). What sets its height?The hard and soft layers of market infrastructure (Arvis et al., 2010;Portugal-Perez and Wilson, 2012). What determines how much is provided? Collective-action and endogenous-policy mechanisms (Olson, 1965;Grossman and Helpman, 1994). Where can policy intervene?Through the lowest tier's levers, each derived from the mechanism above it.Two features carry analytical weight. Each tier constrains the one above it, so the framework reads upward as explanation, downward as intervention. And the feedback path from realised participation back to the governance tier means that as small exporters grow more numerous they form a better-organised constituency, raising the political return to provision. The equilibrium is therefore path-dependent, which is why sequencing and early wins matter more than the volume of reform. The framework's directional claim, that the binding leverage lies in the governance tier because corridors and statutes deliver little if provision is captured by incumbents, rents persist at the border, or implementation stalls at the frontline, is advanced here as a proposition; whether the governance tier in fact binds is an empirical question, and the Discussion sets out the observations that would support or count against it. Figure 1 Fixed and discovery costs of exporting screen small firms out: only the most productive cross the border (Melitz, 2003;Hausmann & Rodrik, 2003) threshold height is set byHard layer transport corridors, multimodal terminals, cold chain, bonded warehousing these interests keep the threshold high (Olson, 1965;Grossman & Helpman, 1994; The framework is a conceptual synthesis of the cited literature; the position of the governance tier as the binding one is a proposition, not an estimated result.If efficient market infrastructure would so plainly help, why is it under-supplied? Not from technical ignorance, but because provision creates winners and losers, and the interests at play work against the dispersed many. Four mechanisms interact, and Table 1 sets out the Uzbekistan-specific form we attribute to each, the reform response it implies, and an indicator by which progress could be judged. Shared assets go unbuilt because they are profitable for no single firm to build (Olson, 1965). Large exporters and the state-owned sector capture the gains of liberalization while dispersed small firms are under-weighted in the policy bargain (Grossman and Helpman, 1994;OECD, 2023). Discretionary customs and permit procedures generate rents that digitization threatens, falling hardest on smaller traders unable to contest arbitrary assessment (Sequeira, 2016;UNECE, 2025). And centrally designed reforms are executed by frontline agencies whose incentives are imperfectly aligned, so high formal implementation rates coexist with firm-level friction (Rasul and Rogger, 2018;Acemoglu and Robinson, 2013). World Bank, 2025). On this evidence we posit that the gains of liberalization accrue disproportionately to concentrated incumbents rather than to dispersed small firms, in line with the logic of Grossman and Helpman (1994).genuinely representative business associations.small exporters per year.Uzbekistan posts one of the region's highest formal trade-facilitation implementation rates, yet firms still report slow and costly day-to-day trading (OECD, 2023;UNECE, 2025). Following Sequeira (2016) (Rasul and Rogger, 2018), though other readings of the same evidence are possible.Measure and publish SME-level clearance outcomes; align frontline agency performance metrics to those outcomes rather than to the count of measures adopted.Median and 90thpercentile clearance time and cost for consignments below a defined value; gap between formal implementation rate and firm-reported friction.Note: entries in the second column combine documented observations, cited in parentheses, with the authors' interpretation of those observations through the corresponding mechanism. The interpretations are propositions consistent with the cited evidence, not established Uzbekistan-specific findings.A reform agenda built on these mechanisms must target the fixed costs screening small firms out and survive the interests preferring the status quo. Consolidation platforms, accredited certification laboratories, cold-chain capacity and e-commerce gateways lower that cost directly. Single-window systems and digital customs should be judged by SME uptake and the real cost of clearing small consignments; the National Trade Facilitation Roadmap for 2025-2030 supplies the framework, but its value will be decided at the frontline (UNECE, 2025).Dispersed firms also need institutions aggregating their interests: an export-promotion agency with an explicit SME mandate, backed by firmer competition enforcement (OECD, 2023).The digital channel deserves particular attention because recent firm-level evidence bears directly on the smallest firms. Digitalization raises the probability that an SME exports, both directly and through productivity gains (Añón Higón and Bonvin, 2024), and online presence is associated with an increase of roughly eight percentage points in the probability of exporting, with the gains extending disproportionately to SMEs and to firms in lower-income countries (Orkoh and Teh, 2025). For a landlocked economy these are among the few margins on which geography imposes no penalty.WTO accession deserves specification rather than assertion. It reduces political rents through three channels. Binding commitments narrow frontline discretion, and border rents are generated principally by discretion rather than protection levels. Transparency and notification obligations make deviations observable, raising the cost of arbitrary treatment. And dispute settlement gives reform-minded agencies an external constraint to invoke against incumbents, converting a distributional fight into a compliance question (Tang and Wei, 2009;OECD, 2025).International experience indicates what is transferable. Firm-level evaluations find that exportpromotion assistance raises exports mainly at the extensive margin, that is, in the number of firms exporting and markets reached, which is precisely the margin at issue here (Volpe Martincus and Carballo, 2008); evidence on African exporters indicates that transit delays and related trade costs, rather than firm capability, account for much of the export shortfall (Freund and Rocha, 2011). The caution is that these gains materialised where procedural reform was implemented at the frontline rather than legislated centrally.The agenda is easier to specify than to execute. Fiscal constraints bind: shared laboratories and cold-chain capacity are capital-intensive, compete with more visible expenditure, and return diffusely and late. Institutional capacity is thin in accreditation, where international recognition requires competence that cannot be legislated. Governance quality conditions everything downstream, since a single-window system administered discretionarily reproduces the rents it was meant to remove. Vested interests resist most sharply where digitization removes discretion from agencies whose staff derive income from it. And coordination failure binds: customs, standards bodies, transport authorities and export-promotion institutions answer to different principals, so incomplete reform can raise compliance costs before lowering them.Sequencing, realistic timelines and compensation for losers belong to the design, not to afterthoughts.The contribution of this Opinion is conceptual: it treats the export selection threshold as an endogenous political outcome and specifies the chain through which provision determines it.The central claims, that market infrastructure is the constraint most amenable to policy and that the governance tier is where the binding leverage lies, are propositions rather than established findings, and we state them as such. Their value is that they are testable. If the governance tier For policy the implication is specific. Progress should be measured by SME-level outcomes:first-time small exporters, median clearance time and cost for small consignments, and the price and waiting time for accredited conformity assessment, not by counts of measures adopted. The export potential of Uzbekistan's small entrepreneurship will not be released by exhortation, training or liberalization alone, but by market infrastructure the state is willing to build and politically protect for firms too small to build it themselves. That threshold is, on the account proposed here, a policy choice, and therefore one that policy can revise.