When global wheat prices surged in 2021 and 2022, consumers across the world braced for a familiar pattern: bread and flour prices climbing in lockstep with the commodity markets. But a new study of Bosnia and Herzegovina’s wheat-to-bread supply chain suggests that the journey from farm gate to bakery shelf is far more complicated than simple pass-through. Using quarterly data spanning a full decade, researchers have produced the first systematic analysis of how price shocks travel through this import-dependent Balkan food market, and their findings reveal a system in which consumer prices respond slowly, weakly, and in some cases asymmetrically to movements in the price of raw wheat.
The research, published in the Journal of Agriculture and Food Research, examines the dynamic relationships between national wheat producer prices and consumer prices of white wheat flour and wheat bread from the first quarter of 2014 through the final quarter of 2024. Bosnia and Herzegovina offers a compelling case study for several reasons. The country produces only about one third of the wheat it consumes, with an average self-sufficiency ratio of 32.35 percent and an import dependency ratio of 83.24 percent over the 2014 to 2023 period. Its wheat sector is fragmented, with half of all farms covering less than two hectares, and its food governance is decentralized, with commodity reserve responsibilities split between the country’s two entities and the Brčko District. This combination of external exposure and institutional complexity makes the market an ideal laboratory for studying how price signals propagate, or fail to propagate, along the food chain.
The analytical challenge lies in distinguishing several distinct economic phenomena that are often conflated in public debate. Cointegration analysis tests whether price series that wander individually nonetheless move together over the long run, sharing a stable equilibrium relationship. Error-correction models then measure how quickly each price returns to that equilibrium after a disturbance. Impulse-response functions trace the timing and persistence of reactions to shocks, while asymmetric specifications ask whether prices rise faster than they fall. Crucially, the authors emphasize, none of these price-based techniques can, on its own, identify market power. Raw wheat represents only one component of the final price of bread, alongside labour, energy, transport, packaging, taxes, and distribution margins, so a weak or delayed response of consumer prices does not automatically signal profiteering anywhere in the chain.
The econometric results are notable for their caution as much as their substance. Conventional Johansen trace tests supported a single long-run relationship linking the three price series, but this conclusion dissolved once the statistics were adjusted for the small sample of just 44 quarterly observations. Under the small-sample adjustment, the null hypothesis of no cointegration could no longer be rejected, meaning the evidence for a stable long-run linkage is suggestive rather than definitive. The authors treat their rank-one vector error-correction model as a conditional baseline, a framework within which further results can be interpreted, rather than as proof of an enduring equilibrium binding wheat, flour, and bread prices together.
Within that conditional framework, one result stands out clearly: the error-correction loading is statistically significant only in the flour consumer-price equation, with an estimated coefficient of 0.4795. This indicates that flour prices are the series most strongly associated with correcting deviations from the estimated long-run relationship. The wheat producer-price equation shows no significant adjustment, and the bread equation only marginal adjustment at the ten percent level. Importantly, because the flour variable is a consumer price that bundles together milling, packaging, wholesale, and retail components, this finding cannot be pinned on millers specifically; it identifies adjustment in an observed aggregate price series, not a mechanism at a particular stage of the supply chain.
The dynamic analysis reinforces the picture of attenuated transmission. Following a one-standard-deviation shock to wheat prices, the flour price response turns positive from the second quarter onward and gradually builds, while the bread response remains considerably smaller throughout a ten-quarter horizon. Yet the 95 percent confidence intervals for both downstream responses include zero at every horizon, meaning the study cannot statistically confirm that flour or bread prices respond significantly to wheat shocks at all. The point estimates are consistent with delayed and weakened pass-through, but the data do not establish it. Forecast-error variance decompositions add a complementary descriptive layer: by horizon ten, wheat innovations account for roughly 45 percent of flour price forecast uncertainty and about 37 percent of bread price uncertainty under the baseline ordering, suggesting that wheat shocks become increasingly relevant for downstream prices over longer horizons even when individual responses lack statistical significance.
The clearest evidence of non-uniform behaviour emerges in the bread market. Heteroskedasticity-robust Wald tests reject the equality of positive and negative error-correction coefficients only in the bread consumer-price equation, indicating that bread prices correct upward and downward deviations from the estimated long-run relationship at statistically different speeds. In the short run, positive lagged wheat price changes carry a significant individual coefficient in the bread equation, but the formal test comparing positive and negative wheat changes reaches only marginal significance at the ten percent level. The authors are careful to note that this asymmetry, while consistent with international evidence of rockets-and-feathers pricing in bakery markets, cannot be attributed to market power, retail conduct, or any specific mechanism without additional data on costs, margins, and concentration.
The study’s context matters for interpreting these patterns. Bosnia and Herzegovina not only imports most of its wheat but also depends heavily on imported flour, bringing in an average of nearly 44,000 tonnes annually from CEFTA countries between 2021 and 2024, while its domestic grain-milling industry contracts at roughly 2.8 percent per year in business numbers. Bread remains a dietary staple, with per capita consumption estimated at about 48 kilograms, and the market is projected to grow. A post-2021 indicator in the bread equation is positive and significant, capturing higher average quarterly bread-price growth during a period that layered pandemic disruptions, the Ukraine-related commodity shock, and energy-price inflation on top of one another. The authors interpret this as a broad shock-period effect rather than isolating any single cause, and structural stability diagnostics detected no permanent break in the bread-price dynamics.
These findings sit within a broader European literature that shows how sensitive price-transmission conclusions are to product, period, and method. Studies of Italian pasta and bread chains, Hungarian wheat and flour markets, Ukrainian wheat-flour linkages, and Czech cereal chains have all documented varying degrees of cointegration, asymmetry, and slow adjustment, with new equilibria sometimes taking years to reach. The Bosnian results echo this heterogeneity while adding a first empirical anchor for a Western Balkan market outside Serbia. For policymakers, the message is less about intervening at a particular stage of the chain and more about building better market information systems. The authors recommend integrating domestic producer prices with import prices, exchange rates, major processing costs, and genuinely stage-specific flour and bread prices, which would allow analysts to separate external cost transmission from other drivers of consumer-price adjustment.
The limitations of the analysis are candidly acknowledged and worth heeding. Forty-four quarterly observations limit statistical power, and aggregating monthly consumer prices to quarterly averages may smooth away short-run responses. The flour and bread variables are consumer prices that cannot identify firm-level margins or specific supply-chain stages, and the model omits potentially important common drivers such as energy costs, exchange rates, and quantities. Relative-price checks provided mixed evidence on whether common inflationary trends explain part of the observed nominal co-movement. Longer, higher-frequency, and stage-specific datasets, the authors argue, would enable more flexible nonlinear models and firmer conclusions. For now, the study delivers a nuanced message for a region still digesting the food-price turbulence of the early 2020s: in Bosnia and Herzegovina’s wheat-to-bread chain, shocks travel slowly and unevenly, bread prices adjust asymmetrically to deviations from equilibrium, and reading market power into every delayed price cut would be a mistake the data cannot support.
Subject of Research: Vertical price transmission in the wheat-flour-bread supply chain of Bosnia and Herzegovina
Article Title: Dynamic vertical price transmission in the wheat–flour–bread market: Evidence from Bosnia and Herzegovina
Article References: Mrdalj, V., Stojisavljević, Đ., Kovachevikj, M. G., Ostojić, A., & Mizik, T. (2026). Dynamic vertical price transmission in the wheat–flour–bread market: Evidence from Bosnia and Herzegovina. Journal of Agriculture and Food Research, 31, Article 103346. https://doi.org/10.1016/j.jafr.2026.103346
Image Credits: AI Generated
DOI: 10.1016/j.jafr.2026.103346
Keywords: price transmission, wheat, flour, bread, Bosnia and Herzegovina, cointegration, vector error-correction model, food security, import dependency, asymmetric adjustment, agricultural economics, food prices
Cite Scienmag News
APA
MLA
Chicago
Alan Morgan. (October 3, 2026). When Wheat Prices Spike, Bread Doesn’t Always Follow: New Evidence from Bosnia and Herzegovina. Scienmag. https://scienmag.com/when-wheat-prices-spike-bread-doesnt-always-follow-new-evidence-from-bosnia-and-herzegovina/
Alan Morgan. “When Wheat Prices Spike, Bread Doesn’t Always Follow: New Evidence from Bosnia and Herzegovina.” Scienmag, 3 October 2026, https://scienmag.com/when-wheat-prices-spike-bread-doesnt-always-follow-new-evidence-from-bosnia-and-herzegovina/. Accessed 3 October 2026.
Alan Morgan. “When Wheat Prices Spike, Bread Doesn’t Always Follow: New Evidence from Bosnia and Herzegovina.” Scienmag. October 3, 2026. https://scienmag.com/when-wheat-prices-spike-bread-doesnt-always-follow-new-evidence-from-bosnia-and-herzegovina/
Copy citation
Download RIS
Tags: agricultural economicsagricultural supply chain in Bosnia and Herzegovinaasymmetric adjustmentasymmetric price response in food marketsBosnia and Herzegovinabreadcointegrationdomestic bread and flour priceseconomic analysis of food price dynamicseffects of international wheat market on local consumer pricesflourfood price transmission delay and weak correlationfood pricesFood securityglobal wheat price shocksimpact of global commodity fluctuations on local food pricesimport dependencyimport dependency in Balkan food marketsprice transmissionsupply chain dynamics in wheat-to-bread processvector error-correction modelwheatwheat price transmission in Bosnia and Herzegovinawheat self-sufficiency and import reliance
