The commune of Sidéradougou, located in the Cascades region and Comoé province of Burkina Faso, was chosen as the study area. This region is representative of the challenges related to insecurity and volatility of livestock markets, while being an important center of pastoral and agricultural activities.

The data come from the West African Centre for Economic and Social Studies (CESAO), which conducted systematic field surveys at the main livestock market of Sidéradougou between June 2021 and April 2022. The unit of observation is a single market day. In Sidéradougou, the livestock market meets on a fixed weekly schedule; CESAO enumerators recorded, for each market day, the number of animals offered for sale and the transaction prices (in FCFA) for cattle, sheep, and goats separately. Monthly averages were then computed from the individual market-day records.

The study covers nine months: June, July, August, September, and October 2021, and January, February, March, and April 2022. The months of November and December 2021 are absent from the dataset because CESAO field operations were suspended during that period due to security constraints, leaving a total of 35 market-day observations per species after aggregation. The observation count is identical across the three species because price and supply were recorded simultaneously on each market day.

A descriptive analysis was carried out, including the calculation of basic summary statistics (means, standard deviations, minima, maxima) to describe attendance and price trends across the nine months covered. Graphical visualizations are also produced to highlight patterns in price and supply evolution over time.

Ordinary least squares (OLS) regression describes only the conditional mean of prices and is sensitive to outliers. Livestock prices in informal markets are known to be heteroscedastic and to respond asymmetrically to supply shocks depending on prevailing price levels: sellers facing distress conditions at low-price periods behave differently from those operating in high-price festive periods. Quantile regression ([9]) addresses this by estimating the full conditional distribution of prices, allowing the effect of supply to differ at the 25th, 50th, and 75th percentiles. The 25th quantile captures low-price market conditions (typically the lean season), the median captures typical market conditions, and the 75th quantile captures high-price conditions (typically festive or post-lean periods). Standard errors are obtained by bootstrap with B = 500 replications, following the recommendation of [8] for small samples.

The OLS benchmark solves:

where y i is the price of species s at observation i , x i is the vector of covariates, and β the coefficient vector.

Quantile regression for quantile q ∈ ( 0 , 1 ) solves ([10]):

where ρ q ( u ) = u ( q − 1 u < 0 ) is the asymmetric absolute-loss function and β q is the quantile-specific coefficient vector.

The three estimated models are:

where D m , i are month dummies ( m = 1 , ⋯ , 9 corresponding to June 2021 as the baseline, July 2021, ..., April 2022). The month dummies absorb the three contextual periods defined in Section 1: months within the lean season (July-September 2021), the holiday period (October 2021, January 2022), and the security-effect period (March-April 2022). This dummy specification is the one actually estimated and reported in Tables 5-7; it is equivalent to the period-indicator specification described conceptually in the Introduction.

Table 1 presents descriptive statistics of cattle prices and supply for the nine months covered. Average prices show moderate fluctuations across the study period. Prices are lowest in September 2021 (mean 141,912 FCFA), a month that falls within the lean season when household food stocks are limited and economic pressure is highest ([13]); the elevated supply in that month (mean 69 animals) is consistent with distress-selling behavior, though this interpretation cannot be confirmed from the price and supply data alone. In January and February 2022, mean prices are among the highest observed (155,833 and 156,750 FCFA respectively), coinciding with the post-lean recovery period. The increase in cattle supply in February and March 2022 (means of 68 and 65, respectively) is followed by a sharp decline in April 2022 (mean 37), a month that coincides with the documented intensification of security incidents in the commune. Whether this decline reflects sellers exiting the market due to insecurity or a post-peak seasonal correction cannot be determined with certainty from the available data.

The sheep data show notable fluctuations in prices and supply across the study period. The highest mean sheep price is observed in January 2022 (48,556 FCFA). A plausible explanation is the proximity to Eid al-Adha (Tabaski), which fell on July 9, 2022, but pre-festival demand typically builds several months in advance as buyers seek to secure animals early. This interpretation is consistent with the pattern, though it cannot be confirmed without buyer-level data. Separately, supply in September 2021 was relatively high (mean 73 animals) while prices were low (mean 37,169 FCFA), a pattern consistent with lean-season supply pressure. These dynamics illustrate how seasonal and cultural calendars may interact to shape market outcomes, though the descriptive data alone cannot establish direction of causation (Table 2).

Table 1. Descriptive statistics of cattle prices and supply (FCFA).

Source: Authors, based on CESAO data.

Table 2. Descriptive statistics of sheep prices and supply (FCFA).

Source: Authors, based on CESAO data.

Goats show marked variations in prices and supply throughout the study period. Goat prices peak in January 2022 (mean 25,667 FCFA), which may reflect heightened festive demand, though the exact mechanism cannot be confirmed from market-level aggregates alone. Supply peaks in September-October 2021 and again in February-March 2022. The October 2021 supply peak (mean 101 animals) may reflect the end of the rainy season, which makes animal management more difficult and could induce early sales, though this remains a plausible conjecture. February and March 2022 supply peaks coincide with the onset of security-affected months in the study area (Table 3).

Table 3. Descriptive statistics of goat prices and supply (FCFA).

Source: Authors, based on CESAO data.

Figure 1 and Figure 2 illustrate the joint evolution of prices and supply over the nine months covered. Three broad patterns emerge. First, in the lean-season months (July-September 2021), supply is relatively high and prices are moderate to low, a pattern consistent with economic pressure on herding households. Second, in October 2021 and January 2022, prices rise while supply contracts, consistent with pre-festive herd retention. Third, from March to April 2022, supply contracts sharply and prices decline, a combination that may reflect security-induced withdrawal from the market rather than a demand-side price drop. These patterns are descriptive; the regression analysis below allows a more systematic assessment of the supply-price relationship controlling for month-specific conditions (Table 4).

Figure 1. Evolution of livestock prices (June 2021-April 2022). Source: Authors, based on CESAO data.

Figure 2. Evolution of livestock supply (June 2021-April 2022). Source: Authors, based on CESAO data.

Table 5 presents the quantile regression results for cattle prices. The supply coefficient is negative across all specifications but reaches statistical significance only at q0.75 (−217.391, p < 0.05). It is not significant in OLS, q0.25, or q0.50. This means that evidence for a supply-price relationship in cattle is confined to the upper end of the price distribution, corresponding to high-price market conditions. One plausible interpretation is that buyers in high-price periods are more sensitive to the volume of animals available, but this remains a conjecture.

Table 4. Summary statistics of livestock prices and supply.

Source: Authors, based on CESAO data.

Table 5. Regression results for cattle price determinants.

***p < 0.01, **p < 0.05, *p < 0.10. Standard errors by bootstrap (B = 500). Source: Authors, based on CESAO data.

The month dummies reveal significant differences relative to the June 2021 baseline. Positive and significant coefficients for October 2021, January, and February 2022 across multiple quantiles indicate that prices in these months were substantially higher, consistent with the pre-festive and post-lean periods defined earlier. Prices in March and April 2022 are not significantly different from the baseline at any quantile, suggesting that the security-affected months did not produce a statistically distinguishable price effect in cattle, possibly because reduced supply and reduced demand cancel each other out. The pseudo R² values range from 0.30 to 0.42, indicating that month effects and supply together explain a meaningful but partial share of price variation.

Table 6 presents the quantile regression results for sheep prices. Unlike cattle, the supply coefficient is negative and statistically significant across all quantiles and in OLS, indicating a robust supply-price relationship for sheep throughout the conditional price distribution. January 2022 stands out with a positive and significant coefficient at q0.25 and q0.75, indicating that sheep prices in that month were markedly higher than in the June 2021 baseline, net of supply effects. January 2022 falls approximately six months before Eid al-Adha (July 9, 2022); while this timing is somewhat early for a direct festival effect, it may reflect early procurement behavior by buyers or general seasonal demand for small ruminants at the start of the year. This interpretation is speculative and would require transaction-level data to confirm. The pseudo R² values range from 0.41 to 0.60.

Table 6. Regression results for sheep price determinants.

***p < 0.01, *p < 0.05, *p < 0.10. Standard errors by bootstrap (B= 500). Source: Authors, based on CESAO data.

Table 7 presents the quantile regression results for goat prices. As with sheep, goat supply has a negative and statistically significant coefficient across all quantiles and in OLS, confirming a robust negative supply-price relationship for this species. The month coefficients are positive and significant across nearly all periods relative to the June 2021 baseline, indicating sustained higher prices throughout the study window net of supply. The March 2022 coefficient is among the largest (OLS: 7,428 FCFA; q0.75: 9,636 FCFA), which is consistent with the security-affected period reducing effective supply while residual demand sustained prices. However, this explanation is speculative given that March 2022 dummies capture all time-varying factors in that month, not security effects in isolation. Pseudo R² values range from 0.44 to 0.71, the highest of the three species models.

Table 7. Regression results for goat price determinants.

***p < 0.01, **p < 0.05, *p < 0.10. Standard errors by bootstrap (B= 500). Source: Authors, based on CESAO data.

The results of this study confirm that livestock markets in the municipality of Sidéradougou exhibit strong seasonal and contextual variation across the June 2021-April 2022 window. The findings should be interpreted as associations rather than causal effects, given that the month-dummy specification controls for average differences across periods but cannot isolate the specific mechanisms driving price changes.

The lean season (July-September 2021) is associated with elevated supply relative to June 2021, while prices remain moderate or low. This pattern is consistent with the economic-pressure hypothesis documented by [13], whereby households sell animals to meet immediate cash needs during food-scarce months. The pre-festive and post-lean months (October 2021, January-February 2022) show higher prices net of supply effects, which may reflect demand increases or supply retention by sellers anticipating better prices, though neither mechanism can be confirmed from aggregate market data alone.

The supply-price relationship differs importantly across species. For sheep and goats, supply has a negative and statistically significant effect on prices at all quantiles, indicating a robust negative relationship across the full conditional distribution. For cattle, the supply effect is significant only at q0.75, meaning the evidence is confined to high-price conditions. This heterogeneity justifies the quantile regression approach: OLS estimates of the cattle supply effect, for example, are negative but not significant, and would have led to an erroneous conclusion of no relationship at higher price levels where the effect is actually substantial.

Security events in March-April 2022 coincide with a decline in supply and mixed price signals. The inability to separately identify a “security effect” from general month variation is a recognized limitation of the current design. Future work incorporating event-level security data alongside market records would permit a more rigorous assessment. These observations are consistent with the documented disruption of herder mobility and market access described in [7] and [16], but they remain descriptively consistent rather than causally established.