Climate Change and Crop Production in Somalia: What a 2025 VECM Study Reveals

A digest of Nur & Sheikh Ali (2025), Research on World Agricultural Economy

Most Somali crops are rain-fed, so rising temperatures, erratic rainfall and recurrent droughts, floods and locust outbreaks bear directly on food security. Several studies have examined individual crops such as sorghum and maize, but fewer have modelled total crop output against climate and farm inputs together. This 2025 study, by economists at the Central Bank of Somalia and SIMAD University, uses national data for 1990–2022 and a vector error correction model (VECM) to estimate how temperature, rainfall, CO₂ emissions, agricultural land and labour relate to crop production in the short and long run. Its tables warrant a closer reading than its headline figures, which makes it a useful case for researchers and policymakers working on climate adaptation and food security.

The study covers 1990–2022, giving 33 annual observations, and uses a VECM with Johansen cointegration and one lag. The crop error-correction term is −0.014 and is not significant, and none of the short-run drivers in the crop equation is significant.

Article at a Glance

The article is "Climate Change and Crop Production Vulnerability in Somalia: A VECM Analysis for Sustainable Agriculture," by Abdulkadir Mohamed Nur (Central Bank of Somalia) and Ali Yassin Sheikh Ali (SIMAD University). It was published in Research on World Agricultural Economy, Vol. 6, No. 3, pp. 355–369 (Nan Yang Academy of Sciences; open access, CC BY-NC 4.0), in 2025 (received 6 March; accepted 15 April; published online 16 July). Its research area is agricultural and climate economics, and its study context is Somalia, using national annual data for 1990–2022 (World Bank, FAO, SESRIC). The method is quantitative time series: ADF and PP unit-root tests, Johansen cointegration, and a vector error correction model. The DOI is https://doi.org/10.36956/rwae.v6ix.1815 (as printed; "v6ix" may be a placeholder, so verify).

The Research Problem

Research from Asia and elsewhere in Africa documents climate damage to crops and the need for adaptation. For Somalia, the authors argue, earlier work has focused on specific crops or short-term climate variability, often without econometric estimates, and has paid little attention to how land and labour interact with long-term changes in temperature, rainfall and CO₂.

Research Objective

The study estimates the short- and long-run effects of climate factors and farm inputs on crop production, to inform adaptation policy. Its overarching question asks how Somalia can avoid, reduce and mitigate climate impacts while sustaining agricultural output and food security. The authors expect temperature and CO₂ emissions to have negative effects, and rainfall, agricultural land and labour to have positive effects.

How the Study Was Conducted

A single-country time-series design with 33 annual observations. Crop production (thousand tonnes) is the dependent variable; average temperature, CO₂ emissions (kilotons), annual rainfall, agricultural land and labour are explanatory. All variables are logged, and the model draws on the Ricardian, Cobb-Douglas and induced-innovation frameworks.

The analysis followed six steps. In step 01, Unit roots, ADF and PP tests led the authors to classify five series as I(1) and labour as I(2). In step 02, Lag length, one lag was selected (SC), while AIC, HQ, FPE and LR favour two. In step 03, the Johansen test, the trace test rejects every rank, and the max-eigenvalue test gives p = .067 for no cointegration. In step 04, Estimation, a VECM with one cointegrating equation was fitted, giving short- and long-run coefficients. In step 05, Normality, the Jarque–Bera test passes for all equations except rainfall (p = .004). In step 06, Autocorrelation, the LM test detects autocorrelation at lag 1 (p = .019) but not at lag 2.

Key Findings

The long-run results come from Table 8 of the article, where the cointegrating vector is normalised on crop production = 1. The "implied elasticity" is Skilful's reading: under this normalisation, the long-run effect is the negative of the reported coefficient, per 1% change. CO₂ emissions have a reported coefficient of 5.786 (p < .001), an implied elasticity of about −5.8%, while the authors' reading is −57%. Average temperature has 300.536 (p < .001), implying about −300%, while the authors' reading is more than −100%. Rainfall has −7.617 (p = .002), implying about +7.6%, while the authors' reading is +70%. Agricultural land has −1.235 (p = .490), which is not significant, while the authors' reading is +12%. Labour has −6.846 (p < .001), implying about +6.8%, while the authors' reading is +68%.

1. Long run: CO₂ and temperature are linked to lower output; rainfall and labour to higher. The signs match the authors' expectations and are significant at 1%, except agricultural land. The magnitudes the authors report, however, are roughly ten times the coefficients in Table 8.

2. Short run: the paper reports gains from CO₂, temperature and rainfall. The authors link these to farmers' traditional adaptation. In Table 7's crop-production equation, however, none of the lagged terms is significant, and CO₂, rainfall, land and labour carry negative signs.

3. Crop output does not significantly adjust toward equilibrium. The crop equation's error-correction term is small and insignificant (−0.014, SE 0.017), so the data do not show crop production returning to the long-run relationship.

Reading note: The headline figures — a 57% fall per 1% rise in CO₂ and over 100% per 1% rise in temperature — do not match the long-run coefficients as printed, and the short-run "4.8%" CO₂ effect does not appear in the crop-production equation. The paper also says the max-eigenvalue test confirms cointegration, although its first row is not significant at 5%. Readers should rely on Tables 6–8.

What the Study Contributes

The contribution is primarily empirical: a national, multi-variable time-series model of Somali crop output that combines climate variables with land and labour over three decades. The authors call for climate-resilient and climate-smart agriculture, better irrigation, technological improvement and targeted support to help farmers manage climate risk.

Important Limitations

Identified by the authors: Data availability is limited, and environmental and socio-economic factors such as soil erosion, market competition and political violence are omitted.

Skilful cautions arising from the study design: 33 observations is a small sample for a six-variable system. The trace test rejects every rank, which implies all series are stationary — at odds with a VECM — and several series test as stationary in levels. A temperature elasticity near −300 is implausible when average temperature ranges only from 26.6 to 27.3 °C. National CO₂ emissions are not atmospheric CO₂; they may proxy economic activity rather than a direct effect on crops. Finally, autocorrelation at lag 1 weakens standard errors.

Why This Research Matters

Somalia's food security depends heavily on rain-fed farming, and the country's climate adaptation planning needs quantitative evidence. The study adds a national time-series perspective, and its data and specification offer a starting point for more robust modelling.

Skilful Research Insight

Editorial commentary — not a finding of the study. Three lessons stand out. First, a cointegrating vector must be translated before it is interpreted: with the dependent variable normalised to one, signs flip and coefficients are elasticities, not tens of percentage points. Second, test magnitudes for plausibility against the data range — an effect that implies output collapsing with a fraction of a degree of warming signals a specification problem. Third, future work could use crop-specific yields, seasonal rainfall for the Gu and Deyr seasons, regional panels and ARDL or Bayesian methods suited to short samples.

Read the Original Research

The original article is "Climate Change and Crop Production Vulnerability in Somalia: A VECM Analysis for Sustainable Agriculture," published in Research on World Agricultural Economy, 6(3), 355–369 (2025), by Abdulkadir Mohamed Nur and Ali Yassin Sheikh Ali. Its DOI is https://doi.org/10.36956/rwae.v6ix.1815 (verify), and the publisher/journal page is NASS, journals.nasspublishing.com/index.php/rwae (CC BY-NC 4.0). Skilful summarises; all findings belong to the original authors.

SEO Information

The SEO title is "Climate Change and Crop Production in Somalia: 2025 VECM Study Summary." The meta description is "A Skilful digest of a 2025 VECM study on how temperature, rainfall, CO₂, land and labour relate to Somalia's crop production, 1990–2022." The URL slug is /climate-change-crop-production-somalia-vecm. The primary keyword is climate change and crop production in Somalia. The secondary keywords are food security Somalia; VECM; Johansen cointegration; rainfall and temperature; climate-smart agriculture.