The official promotional discourse that has persisted over the past month regarding the Azerbaijani government’s recently approved Agricultural Development Program for 2023–2030 is reminiscent of the grandiose reports devoted to five-year plans during the Soviet era. The document is presented as though no fewer than ten state programs and road maps concerning national food security and the development of the agricultural sector have been adopted over the past twenty-five years, and as though the government had, for the first time, produced a policy document capable of bringing progress to the sector. Yet, in the relatively recent past, the same government adopted a comprehensive road map for agricultural development, as well as a range of state programs aimed at increasing food production and promoting the development of viticulture, cotton growing, tobacco cultivation, rice farming, tea production, citrus cultivation, and sericulture. Unfortunately, despite their strategic significance, these numerous government policy documents have failed to reverse the decline of the agricultural sector. Reports issued by the State Statistical Committee indicate that growth in the sector has slowed substantially, particularly over the past two years, and effectively came to a standstill last year.
Program documents are not magic wands. Without competent implementation, even the most sophisticated strategy cannot halt a crisis or deliver the progress it promises. But implementation is only part of the problem. A program may itself be fundamentally flawed—especially when its targets are detached from reality and its performance indicators fail to capture the depth of the crisis in the sectors they are supposed to address.
This is precisely where the Azerbaijani government’s agricultural policy falls short. The problem is not merely the absence of effective governance in the sector; the new five-year program is itself riddled with serious omissions and structural weaknesses. This article examines one particularly revealing area: the program’s approach to livestock farming and the major problems embedded in its targets and priorities.
The Current State of Azerbaijan’s Livestock Sector
Official agricultural statistics paint a deeply contradictory picture of Azerbaijan’s livestock sector. Given the notable shifts in livestock numbers since 2008, the period from 2008 to 2025 is used as the basis for analysis. Over these years, the cattle population declined by approximately 40,000 head, falling to around 2,4 million. The number of sheep and goats dropped by nearly 15%—equivalent to about 1,2 million head—bringing the total down to 6,7 million. Poultry was the only category to register substantial growth, increasing by more than 40% to reach 29 million birds.
The most striking aspect of the official data is that meat production reportedly increased by 71%—equivalent to 165.000 tons—to more than 390.000 tons, even as the number of cattle, sheep, and goats declined and only the poultry population expanded. Such a sharp rise in slaughter output during a period of falling cattle and small-ruminant numbers inevitably raises questions. This is particularly significant given that roughly two-thirds of Azerbaijan’s meat production—around 250.000 tons—is supplied by cattle, sheep, and goats. Meat production can, of course, rise even as livestock numbers decline. Improvements in the genetic quality of the herd could be one explanation. Yet the program’s own figures show that pedigree cattle account for only 2,5% of Azerbaijan’s total cattle population. This is clearly too small a share to plausibly explain an increase of more than 70% in meat output.
Other explanations may be offered. Lower rates of premature livestock mortality, an increase in the number of animals sent to slaughter, slaughtering animals at higher live weights, and a rise in the number of offspring produced per 100 breeding females are all important factors that can improve productivity. These factors could indeed help expand meat production even without an increase in livestock numbers. Yet no official statistics are available for any of them. If the authorities are so eager to present a 71% rise in meat production over seventeen years as a major achievement, why have they not considered it necessary to include these basic productivity indicators in the country’s agricultural statistics?
The answer is difficult to ignore: Once a comprehensive and internally consistent set of indicators is made publicly available, the scope for manipulating headline growth figures becomes far more limited. Why, for example, are productivity indicators routinely published for crop farming—such as grain, cotton and vegetable yields per hectare—while no official data are provided on the average meat yield of an animal or bird sent to slaughter?
One further detail must be added to this already highly contradictory statistical picture: During the same period in which domestic meat production reportedly increased by more than 70%, Azerbaijan’s meat imports more than tripled, approaching 90.000 tons. In 2008, imports accounted for roughly 13% of the country’s total meat supply; today, that share is close to 25 per cent.
The agricultural development program also claims that agro-parks and intensive farming enterprises will play a major role in driving sectoral growth. Yet official statistics make little effort to provide any meaningful data on their performance. How many livestock are held by agro-parks and agricultural enterprises more broadly? What share of the country’s meat, milk, grain, and fodder-crop production do they account for? These basic questions remain unanswered.
At first glance, weak statistical reporting may not appear to be a critical problem. In reality, however, it is impossible to design credible, high-quality strategic policy documents without reliable data capable of revealing conditions on the ground. Nor can sound management decisions be made in the absence of accurate evidence. The newly approved agricultural development program itself clearly reflects this severe deficiency in the country’s statistical system.
Vague Targets for a Livestock Sector in Crisis
The program applies an inconsistent approach to setting its targets. For example, it aims to raise the share of pedigree cattle in the total herd from 2,5% to 10%—a clear and measurable indicator. Yet while it also states that the number of sheep and goats in agro-parks and semi-intensive farms will increase by 670.000 head by 2030, it provides no baseline figure for 2025. Without a starting point, the target cannot be meaningfully assessed or independently monitored. Could the number at the program’s starting point perhaps already be 665.000 head? If so, how can an increase of just 5.000 animals over the program’s five-year implementation period be presented as an indicator of progress? Target indicators defined without baseline data are inherently weak, especially in a state program. The document, for instance, promises to increase beef production by 20 per cent within five years, yet says nothing about how the average slaughter weight of cattle, a basic measure of productivity, is expected to change.
The program also projects a sharp increase in meat production over a relatively short period, but provides no target for reducing the country’s dependence on meat imports. This omission is particularly revealing. Most of the proposed measures—improving herd genetics, implementing selection and breeding programs, and expanding semi-intensive farms—are explicitly intended to raise productivity. Yet the program’s own outcome indicators conspicuously avoid measuring productivity at all.
The government declares that the livestock sector’s dependence on imported feed will be reduced, yet provides no information on the current level of that dependence or the target level to be reached by 2030. Some of the measures presented as outcome indicators—such as establishing a digital system for monitoring animal diseases—are not indicators at all. They are merely activities intended to contribute to an outcome, rather than measurable results against which the program’s success can be assessed. The document fails to establish any clear causal link between proposed activities—such as improving the feed base, introducing electronic livestock registration, and providing subsidies—measurable outcomes, including higher milk and meat yields per animal, lower mortality rates, and improved calving rates, and broader impacts such as reduced dependence on meat imports and increased farmer incomes. Without such a results chain, it is impossible to determine how the proposed interventions are expected to produce the program’s stated objectives.
Finally, although the analysis presented here exposes the severe limitations of the agricultural sector’s statistical database, the new program makes no provision whatsoever for improving the sector’s statistical reporting system.
