Behind 5.2 Trillion Yuan in Risk Coverage: The Truth of Agricultural Insurance Claims Settlement

Yi Fujin: The 155 billion yuan is premium income, that is, premiums paid by farm households plus premium subsidies provided by governments at all levels, totalling 155 billion yuan. The 5.2 trillion yuan risk coverage level is based on the assumption that all the subject matter of insurance is damaged and that claims are paid up to the actual sum insured; in that case, it would amount to 5.2 trillion yuan, accounting for 60% of the country’s agricultural GDP, which is already very high. However, the 5.2 trillion yuan is not liquid money. It relies on premiums accumulated over many years before it can be paid out in one go, and it is a rather notional figure; in practice, it is highly unlikely that such a large amount of claims would be paid out all at once. What really matters is the 155 billion yuan. The scale of agricultural insurance premiums is now the largest in the world. Of the 155 billion yuan, about 80% comes from the government, and the remaining roughly 20% comes from premiums paid by farmers.

Yi Fujin: The structure of agricultural production in our country has a very distinctive feature: a large country with smallholders. Currently, about 97% of China’s agricultural producers are smallholders. This fundamental reality means that agricultural insurance cannot be implemented with 100% precision. In addition, given the complexity of agriculture itself, it is also difficult for agricultural insurance to provide personalised, household-by-household, case-by-case services in the way that motor insurance and property insurance do.

Last year, during our fieldwork in Henan, we came across a very typical township with over 20,000 mu of arable land and more than 5,000 farm households. Yet insurance companies are already very short-staffed at the local level. Generally, one insurance company’s local staff member is responsible for two or three such townships, and in some places even more. In other words, one person may have to cover 40,000–50,000 mu of land and over 10,000 farm households. The difficulty is further compounded by the fact that agricultural production is extremely time-sensitive, which also means that the state places very high demands on the timeliness of agricultural insurance. The basic requirement is that the entire process—from loss survey and loss assessment through to the disbursement of claims settlement funds to households—be completed within 14 days.

But the problem is that the loss survey, loss assessment, confirmation and disbursement for tens of thousands of mu of land must be completed within a dozen or so days, creating enormous pressure. It is easy to imagine that, in practice, when faced with large numbers of smallholders, it is almost impossible for insurance companies to carry out loss assessment household by household and plot by plot; this is the greatest obstacle to the implementation of agricultural insurance under the national reality of a large country with smallholders.

As a result, in some areas an alternative model is now being used in which the village collective signs a “collective insurance policy”. Once smallholders in the village are affected by a disaster, the insurance company randomly selects 3–5 plots within the village within 14 days, and on that basis makes a comprehensive assessment of the damaged area and the severity of damage across the whole village. Based on the average level of damage, it confirms the total compensation amount for the village. Once the amount is confirmed, village committee officials—who are in fact insurance assistance staff hired by the insurance company—distribute it according to each household’s actual circumstances, completing claims settlement for each smallholder.

This gives rise to two possibilities.

The first is that, although the current model is nominally called “yield insurance” or “full-cost insurance”, the principle governing its operation is in fact area yield insurance, which provides cover based on the average yield of the whole village. For example, if a farm household’s plot on tidal flats is washed away after a flood, the village as a whole may be unaffected. Under the principle of random sampling, that plot may not be selected, or it may have no material effect on the village’s overall yield.

The second is that, following agricultural land reform, many farm households’ arable land is not in one continuous piece. A farm household with 10 mu of land may be divided into four or five plots, and the land is even more fragmented in mountainous areas. Of the 10 mu, perhaps only one plot is affected by a disaster, and under the terms of the insurance contract it may not necessarily be judged to have been affected by a disaster, so the most basic payout cannot be triggered. But farm households do not accept this; they believe that as long as this plot has been affected by a disaster, they should receive compensation, regardless of the size of the loss. Because farmers have misconceptions about the basic principles of agricultural insurance, and because there has been insufficient publicity and explanation, farm households also have considerable complaints about agricultural insurance.

So, whether in the design of the mechanism or in the inherent characteristics of agricultural insurance, the result is so-called erroneous and missed payouts.

For large-scale farming households, the scale is generally at least 50 mu. In some places, such as the north-east, it may be 100 or 300 mu. Each province basically requires insurance companies to handle such large-scale farming households through separate policy issuance and by signing separate contracts. After a disaster, insurance company staff must specifically go to large farm households to carry out loss assessment. By comparison, the accuracy of agricultural insurance is much higher for this group of farming households.

◉ In 2025, Shaanxi was hit by a severe drought; wheat on the river flats south of the Wei River had turned yellow and withered, and even the flag leaf had dried up; the soil around the wheat roots was scorching to the touch by day. Image source: Kong Lingyu

Yi Fujin: There is another reason why farmers feel that the payout falls far short of their expectations: the payouts under agricultural insurance (full-cost insurance, or materialised cost insurance) actually cover only factor inputs.

The total cost of agricultural production is not like that of industrial production, where all inputs are put in at once; it is incurred gradually. For example, at the seedling stage, only about 50% of the cost may have been incurred; by the jointing stage, cumulative input may reach about 70%; by the flowering stage, about 90%; and only at maturity and harvest is the full cost fully incurred. The payout therefore also depends on the stage of production at which the farmer is affected by a disaster.

Take a straightforward example. Suppose a disaster occurs at the wheat seedling stage—for instance, a late spring cold snap after the New Year causes the wheat to freeze to death at the seedling stage. In that case, the insurance company would normally not pay out up to the maximum coverage level, but would calculate the payout based on the inputs already incurred at the seedling stage. On a total-cost basis, the seedling stage may account for only 50%. So even if the agricultural insurance contract states a coverage level of 1,000 yuan per mu, the crop being frozen to death at the seedling stage may result in a payout of at most 50%, that is 500 yuan per mu, because later costs at the jointing and flowering stages have not yet been incurred. If the disaster occurs at the harvest stage—for example, lodging or sustained heavy rainfall causes a total crop failure—the payout may be 100%. Agricultural insurance contains many technical details, but they are not set out clearly enough in the policy terms. In practice, however, actual claims settlement is carried out in exactly the way described above.

Yi Fujin: This practice is known in the industry as a “premium refund”. In years when farm households are not affected by a disaster, insurance companies refund premiums to them to maintain their willingness to take out insurance in the following year; if a disaster does occur, claims settlement is still carried out according to the normal process.

How should we understand the premium refund issue, and what are the reasons behind it? Because the state’s payouts to large numbers of smallholders are based on loss assessment through sampling and are inherently difficult to get exactly right, and the premiums are set too high, to the point where one could say this kind of insurance is a “defective product”. To maintain farmers’ willingness to take out insurance, insurance companies must give farmers some form of reassurance through agricultural insurance that is not in fact worth this price. In practice, therefore, when there is no disaster, insurance companies return a portion of the premiums to farm households.

But this is by no means a good thing for agricultural insurance, because agricultural insurance ultimately aims to accumulate an insurance fund surplus in years without disaster, so that when disaster actually strikes, it can pay out. In current practice, money that should have formed part of the insurance fund surplus is being paid out, so that when a major disaster actually occurs, there will be no money left to pay claims, creating the so-called problem of “big payouts for minor disasters and small payouts for major disasters”. This is why the 2024 State No. 1 document put forward “dual precision”—precise insurance enrolment and precise claims settlement. This issue has become the most important real-world challenge in agricultural insurance today.

◉ Rice fields submerged by floods show widespread lodging, with mud and water still clinging to the rice panicles and stalks. Image source: Yvonne

Yi Fujin: Insurance companies themselves vary greatly. For example, the largest insurance company in Henan is Zhongyuan Agricultural Insurance, which has its own sub-branches at the grassroots level, namely county sub-branches. In the case of very minor disasters, as long as the amount to be paid is not especially large, the payment can be made through a county sub-branch or a prefecture-level central sub-branch. But in the case of exceptionally large disasters, especially widespread disasters such as the situation last year, a county-level company, or a prefecture-level central sub-branch, or even a provincial company, simply cannot afford to meet the claims.

Because the premium charged is only 40 yuan per mu, and that 40 yuan is meant to cover a payout of 200 yuan if a farm household is affected by a disaster in that year. In other words, the premium collected is insufficient to support a 200-yuan payout. In this situation, it is necessary to rely on risk reserves accumulated in the past—that is, a risk surplus—to support payouts in the event of a major disaster.

In normal years, insurance companies set aside “catastrophe reserves” from premiums. They also buy agricultural reinsurance from reinsurers. If payouts arise that a local company cannot meet, they can be paid from the catastrophe risk reserve, or the risk can be allocated to reinsurers. But a major challenge now is that China has yet to establish a highly developed catastrophe risk response system.

Ideally, for a company such as PICC, if a local branch experiences excess of loss, headquarters can in theory mobilise resources from across the regions to coordinate. But for local agricultural insurance companies such as Zhongyuan Agricultural Insurance in Henan or Guoyuan Agricultural Insurance in Anhui, the position is much more difficult. If a major disaster strikes locally, the pressure to meet claims becomes extremely high, making it hard to spread risk nationwide. This can lead to a situation similar to a bank run, where the company is unable to pay out such a large amount all at once and may face enormous financial pressure.

In the United States, Canada, Japan, and even India, the central government usually establishes an emergency fund, because the fiscal budget already contains such an emergency reserve that can be used, and government bonds can also be issued to mitigate the impact of catastrophe risk. In China, however, the mechanisms in this area are still relatively weak at present. In 2020, at the national level, the China Agricultural Reinsurance Company was established. At present it still operates through 20% reinsurance—for example, if an insurance company collects 1 billion yuan in premiums in one province, it must place 20% of that, that is 200 million yuan in premiums and the corresponding risk, with reinsurers—the China Agricultural Reinsurance Company and local insurance companies have thus formed a risk-sharing mechanism.

But this does not fundamentally solve the problem. Because excess-of-loss risk is widespread—a system-wide catastrophe risk of the kind seen in Henan, Anhui, Shanxi and Shandong—the reinsurance mechanism is certainly insufficient. In such circumstances, the operations of any commercial insurance company are unsustainable. But our country’s institutional development in this area is still relatively lagging.

◉ After the rain, a tractor is stuck in the mud, unable to move. Operating heavy machinery on arable land after rain also compacts the soil and damages the arable land. Image source: Xiaoliushu Farm

Yi Fujin: It sounds quite profitable, and it certainly is. This is because the overall risk across China’s major agricultural products—that is, staple crops—is relatively manageable. Although disasters have intensified year after year, the level of food security has continued to improve, with output rising and harvests being good year in, year out, which means that, overall, disasters in farmers’ food production are broadly manageable. Against this backdrop, smallholders’ collective insurance enrolment and collective claims settlement have saved insurance companies substantial costs. As a result, for a considerable period, the average profit margin in agricultural insurance stood at around 9%.

The issue is not that profit margins are too high, nor is it that insurance companies should not earn this money, but what this money actually is. On paper, of course, it is recorded as profit. But given that agricultural insurance itself is subject to enormous year-to-year differences in the frequency and intensity of disasters, the situation is different. Take the United States, where agricultural insurance is the most developed and has the longest history, for example: in extreme years, the loss ratio can reach 160%, 170%, or even 200%. This would be even more unimaginable in the motor insurance industry, but in agricultural insurance this kind of problem can arise. When a major disaster strikes—for example, in 2022, when there was no rainfall for over half the year in the regions south of the Yangtze River—if payouts were truly made on the basis of actual losses, the loss ratio would be no lower than 200%, which is quite normal.

So, for many insurance companies, this surplus on the books can be described both as profit and as risk reserves used to respond to future catastrophe risk.

But at present, a catastrophe risk protection system in the form of risk coverage has not yet been put in place, and the government is still well behind in formulating the relevant policies. In addition, this money is in effect the benefit that insurance companies receive from their cooperation with the government. Through cooperation with the government, the government bears the costs, while the profits end up belonging to the insurance companies. When this cost surplus is reflected as on-the-books profit for the insurance companies, it should, in the future, be properly pooled at the national and provincial levels in the form of a catastrophe risk reserve or a management fund. In terms of the accounting treatment of insurance companies, a corresponding adjustment process will also be needed in the future.

◉ In 2023, the Guanzhong region of Shaanxi experienced 10 consecutive days of “harvest-spoiling rain”, with some of the wheat suffering moulding. Image source: Lüwo Farm

Yi Fujin: The profit margin may not be that low, but overall it remains relatively high. That said, a number of factors have contributed to this.

First, with climate change, extreme weather events are becoming increasingly frequent, and the pressure on insurance companies to meet claims is indeed growing.

Second, the government has also observed that the average profit margin of insurance companies is indeed very high, because these are public funds: local governments put up real money as premium subsidies. Especially in areas where understanding of insurance is not deep enough, they will demand that, for example, if the finance department allocates 50 million yuan, at least that amount must be spent by year-end. For local governments, this is a good thing, because the money takes effect in the current year and is paid out to farmers. But when a major disaster actually strikes, problems arise.

Third, farmers’ willingness to assert their rights is growing, which has placed significant pressure on insurance companies’ operational compliance. Of course, there may also be moral hazard here. When we conducted research in Henan and Anhui last year, we encountered this problem: some farmers felt that a payout of 80 or 90 yuan was not enough and demanded at least 500 or 600 yuan, because the cap is 1,000 yuan. But, given the claims-paying capacity of insurance companies, it is simply impossible to pay.

Fourth, in recent years, under increasingly strict regulation, insurance companies have also paid out as much as possible.

These four factors together are causing the decline in profit margin that we are seeing now. Is the decline in profit margin a good thing or a bad one? Personally, I do not think it is necessarily a good thing. The catastrophe risk management fund has not yet been established, so if a major disaster strikes, an insurance company’s claims-paying capacity is severely weakened. So, of course, from farmers’ perspective, the decline in profit margin is a good thing, but it is also a danger signal.

Yi Fujin: I do not have a complete grasp of this data, but from what I have gathered indirectly, it is not particularly high, and may even have declined. In some countries, such as Japan, its risk funds have accumulated to 1.6 times its premium income to respond to extreme disasters. But our country is far from reaching this level.

◉ In July 2025, extreme rainfall once again struck Beijing. At Sohu Farm, beside the Qingshui River in Beizhuang Town, Miyun District, all 450 mu of the farm’s land was swallowed by the flood. The flood almost erased every trace of the farm, leaving only a washed-out roadbed, mud submerged in water, and a small amount of construction debris. Image source: Zhen Rui

Yi Fujin: It was indeed entirely commercialised from 1982 through to 2004. The logic of full marketisation failed. This is not a case specific to China; around the world, no government has been able to get agricultural insurance up and running, largely because agricultural risk is too high. On the supply side, the cost of agricultural insurance is also too high. On some occasions, I have half-jokingly said that although agricultural insurance is provided by commercial institutions, its coverage in every part of the country is second only to the national government, and in effect every township has a dedicated insurance company employee. It has to bear the costs of institutional staff, travel, and technical means, and these costs add up to a very high total. The high costs also make agricultural insurance a poor business. Overall, the marketisation of agricultural insurance, without government support, whether fiscal or administrative, has produced no successful cases anywhere in the world and is not feasible in practice.

Yi Fujin: There is also an element of speculation on the part of farmers. The poor state of local finances at present has prevented insurance from playing its proper role in risk coverage, mainly because of the particular logic behind our country’s fiscal subsidies.

The logic behind our current subsidies is as follows: after the insurance company collects 20% of the premium from farmers, the county-level government provides 5% to 10% of the premium as matching funds. Once these matching funds are in place, fiscal subsidies are then provided at the prefecture-level city level, then at the provincial level, and finally at the Ministry of Finance at the central level. The central fiscal subsidy can only be disbursed after it has been confirmed that the subsidies from the three lower levels of government have all been put in place. This has given rise to the so-called premiums receivable problem. In particular, in the central and western regions, local finances are relatively tight, and local governments cannot come up with the money. So, starting from the source of insurance enrolment, although farmers may have already paid, if the local government cannot pay out, the whole insurance chain breaks down and enrolment has not been completed at all. This is currently a very big problem. When local finances are tight, if insurance has not even completed the contract-signing process, how can there be any talk of claims settlement?

Recently, we saw a data point: the central government has actually earmarked roughly 50 to 60 billion yuan in its budget. But in fact, since 2024, the completion rate for central fiscal subsidies has failed to reach 100%. The most important reason is that local governments cannot come up with the money at the source to help farmers pay their premiums, so central funds cannot be put in place. This is currently a particularly pressing problem.

But from the standpoint of insurance implementation, local governments remain crucial; without the endorsement of township governments or village collective organisations, smallholders would simply be unable to participate in agricultural insurance schemes.

◉ In early August 2023, Wuchang in Heilongjiang Province was hit by rare torrential rain and floods, and around 40% of its 2.5 million mu of farmland was submerged. The area shown in the image was originally a paddy field; it was completely submerged by the flood. Image source: Yvonne

Yi Fujin: When relatively minor flooding or concentrated rainfall occurs, insurance companies can indeed respond by providing farmers with items such as water pumps and drought-resistant varieties. But when extreme disasters strike, these measures are certainly of no avail.

So, from a logical standpoint, I do not object to this view. Pre-event preventive action is very important and can indeed be effective against some disasters. But for more serious disasters, pre-event preventive action is in fact of little use.

The most fundamental question is: what exactly is the positioning of agricultural insurance?

If it is positioned as pre-event preventive action, current financial management regulations do not permit it either. From conversations with some friends in the industry, I have learned that 20% of the insurance claims funds or premium funds is management expense, and only part of that 20% can be set aside as a prevention fund; the remaining 80% of premiums cannot be used for prevention.

In any case, if one wants to use agricultural insurance to replace more effective measures such as the construction of farmland water conservancy facilities, upgrading production technology, and the development of high-standard farmland, one is clearly putting the cart before the horse. That is why the division of responsibilities and the positioning of agricultural insurance within China’s overall agricultural policy framework are so crucial.

I have heard many different views. The vast majority of scholars, government officials and insurance company professionals would hold that the government’s premium subsidies for agricultural insurance are, in fact, support for agriculture or for farmers to manage risk. By conventional logic, that is indeed the case.

But there is another problem. When agriculture, and particularly grain crop production, is viewed from a higher perspective, an ample supply of grain—food security, in other words—is, for society as a whole, the cornerstone of social stability and development. It is a public good. Everyone can benefit from food security. In normal years of good harvest, consumers can buy enough grain in the market and enjoy the benefits that food security brings. But when grain output falls, the state uses macro-level regulation, releasing grain from state reserves to stabilise prices and using imports to adjust domestic grain supply. Consumers suffer little loss, but farmers, because of food security as a national strategy, bear the losses, and no one pays farmers a corresponding price premium, or at least some compensation.

In other words, in the course of achieving the strategic goal of food security, society as a whole has enjoyed the benefits of an ample supply of grain, but we have not paid the corresponding risk premium for the risk of reduced grain output.

So from this perspective, the government’s premium subsidy for agricultural insurance is not a subsidy, but the risk premium that the government should pay to achieve food security.

Looking back over more than a thousand years of Chinese history, the so-called harm to farmers from low grain prices—when output rises, farmers do not benefit because prices fall; yet when output declines, farmers also suffer losses due to the state’s effective grain-price regulation. Considering these two matters together, farmers actually bear all the losses, while society shares in the market returns; this is inequitable. In other words, returns and risks are wholly mismatched. The government should therefore take measures to help farmers manage risk, so that low risk can be matched with low returns. From the perspectives of risk management and social equity, the government should absolutely pay this sum.

▼

Click the image to read related articles

▼