Handmade milk tofu really is tastier—it’s not just a rip-off!
Those who have visited the grasslands of Inner Mongolia will no doubt be familiar with steaming salty milk tea and the distinctive dairy delicacies of the Mongolian people. Last summer, while conducting fieldwork in Sanggen Dalai Town, Zhengblue Banner, Xilin Gol League, Inner Mongolia, I was frequently hosted by local herding families who welcomed me with these treats.
At first, I found the pungent smell of fermented dairy difficult to accept. But after a month, milk tea became a staple of every meal, dairy delicacies became indispensable snacks, and they evolved into a key clue for my understanding of the transformations within the local pastoral regions.
Since 2018, the Inner Mongolian government has implemented a ‘Dairy Revitalisation’ policy, with the Xilin Gol League vigorously promoting the development of the dairy industry and providing extensive support to traditional producers. Despite this trend, producers still face numerous hardships; many small-scale producers have not benefited and are gradually exiting the industry.
Why is this? In the eyes of the herders, this is linked to the replacement of production machinery, the loss of traditional skills, and increasingly standardised food regulations—and is further a consequence of the decline of traditional grassland animal husbandry.
I. Dairy Production as Local Knowledge
Much like any skilled artisan, herders proficient in making these delicacies may be unable to specify the exact duration of fermentation, the precise heat required to drain the whey, or the exact timing and force of stirring. Yet, through the repetition of daily labour, they have acquired a form of empirical and operational knowledge—as is the case with 28-year-old Xin Jile.
Xin Jile learned the craft through immersion from a young age and, after graduating from university, followed in his father’s footsteps to run a dairy business.
He believes there is no fixed recipe; it depends primarily on the state of the work and the quality of the fresh milk. If some milk is left over from making milk tofu today, it is allowed to ferment naturally and then mixed with fresh milk the following day to continue production.
The critical step in transforming milk from curd to milk tofu is ‘chu jin’ (developing the stretch)—heating and stirring the mixture until it reaches a resilient, soft state similar to gluten, capable of being stretched into strings.
Once, while heating the raw milk residue, Xin Jile found it impossible to achieve this stretch; the separated acidic whey was simply reabsorbed into the curd during vigorous stirring. He judged that the milk contained too much fat, so he scooped out half of the mixture and added a bucket of frozen, semi-drained raw milk residue. This time, he stirred more carefully, constantly feeling and kneading the mixture with his hands, and the milk tofu soon took shape.
Another producer, Subu, maintains the consistency of the output by adding fresh milk to the raw milk residue: “If I see it’s not stretching, or if the texture doesn’t feel right, I add a bit of milk to ‘bind’ it together.”
For most producers, as long as the milk quality is up to scratch, the process rarely fails. They are confident that they can control production through subtle adjustments based on their experience and skill to ensure a consistent finished product.
This flexible process also gives rise to a rich variety of fermented flavours: Subu’s milk tofu is golden on the surface with a mellow taste, while Xin Jile’s is chewier and can melt into strings when heated.
II. Standardised Production: Can Quantity and Quality Coexist?
Simultaneously, the government vigorously promoted the standardised upgrading of small workshops. Upgraded standardised plants must include six basic areas, such as changing and disinfection rooms, milk purification workshops, and fermentation workshops, utilising electric processing equipment and replacing all production tools with food-grade plastic and stainless steel utensils.
Gerile, who is in his forties and has run a dairy business in Sanggen Dalai for over twenty years, has witnessed the local industry’s evolution from its inception to the modern era. For Gerile, the use of automated machinery has indeed liberated him from arduous physical labour and significantly increased the scale of production.
Before 2020, Gerile used the back room as his workshop, employing a firewood stove, a large iron pot, wooden stirrers, and wooden moulds to hand-craft milk tofu; an hour of stirring would yield less than five kilograms of the product.
In 2020, encouraged by government policy, Gerile opened a dairy processing workshop and invested in automated machinery. Raw milk is now transported to the production workshop via cooling stirring tanks and delivery pipelines. The fermentation workshop uses air conditioning to control the ambient temperature, ensuring the stability of the fermentation process, allowing for the production of at least 15kg of milk tofu per hour.
While output has increased, Gerile has always felt that machine-made dairy products lack a certain something: “Everything is now done in a sealed room with air conditioning to regulate the temperature, so the cream doesn’t have that rich, concentrated aroma.”
Unlike factory fermentation controlled by air conditioning, traditional fermentation requires the milk to be ventilated and exposed to the air—fresh milk is left untreated to react with microorganisms in the environment. After natural fermentation for 12 to 72 hours, the milk transforms into yoghurt, and the separated fats and whey are further processed into dairy delicacies. In Gerile’s view, the rich aroma of these products depends heavily on whether they can come into contact with natural bacterial flora during fermentation.
“The technology used to make dairy products with machines is different from the craft we used at home, so the taste is bound to be different. As mechanisation increases in the future, traditional dairy culture will slowly vanish,” Gerile worries.
The equipment used in standardised workshops has also caused subtle changes in the flavour of the dairy products.
Traditional wooden tools were selected and crafted by the herders themselves; the milk tofu moulds, often carved with ornaments and Mongolian script, are particularly exquisite. Wood is not only aesthetically pleasing but can also absorb lactose, facilitating a “secondary fermentation” of the product. In many regions, wooden vats remain the preferred choice for fermenting yoghurt and churning butter.
Herders have expressed reservations about this. Natural fermentation is, by its very nature, a product of an open environment. Overly sterilised equipment severs the organic connection with the ingredients, turning dairy products into assembly-line commodities and stripping them of their original unique flavour and cultural significance.
III. Upgraded, and then what?
Although local governments have introduced various incentive policies, in practice, only producers who have already accumulated a certain amount of capital can access the benefits of the Dairy Industry Revitalisation Plan and afford the costs of mechanised production and standardised upgrading.
Currently, there are three standardised workshops in Sangengdalai, while more than ten other dairy shops remain family-run handicraft workshops. Most of these small family businesses find the cost of further expanding their production scale prohibitive.
Sub, who runs a dairy shop in Blue Banner, bluntly admits that he lacks the funds to buy machinery or hire staff, meaning his production scale can only remain as it is.
At the end of 2022, Master Lyu, who had run a dairy shop for seven years, even closed his business: “Dairy products are becoming more expensive, but business is getting harder. Many people have started mechanising; they use milk from large ranches, each batch comes with a quality inspection report, they produce in huge volumes, and their sales channels are wider. I just can’t compete with them.”
The few operators who have kept pace with standardised upgrades were mostly people who had previously worked in other professions and had some savings.
It cost Gerile 600,000 yuan to transform a family handicraft workshop into a standardised small workshop, and that was only for hardware investment in machinery and facility renovation: an electric heating stirring boiler costs about 20,000 yuan per unit, a heating centrifuge and an electric heating rack for milk skin cost about 10,000 yuan each, and the renovation of the 600-square-metre workshop cost approximately 500,000 yuan. For most herders living on the brink of poverty, this is an almost astronomical sum.
Mechanised production not only implies a substantial fixed investment, but it also challenges the local herders’ perception of what constitutes a high-quality dairy product.
From the herders’ perspective, the quality of the milk source is the fundamental factor determining the quality of the final product: “Only dairy products made with local milk have this specific taste. Local cows’ milk has a higher fat content; the milk tofu produced from it has a yellow layer of oil on the outside—that’s the kind that tastes best.”
Their reasoning is sound. Local Mongolian yellow cattle spend most of their time grazing on open pastures, have long lactation cycles, and produce high-concentration milk with excellent quality, resulting in dairy products with a higher fat content.
The flavour of milk varies according to the breed of cow, the composition of the feed, and the change of seasons. Herders, however, hold an inherent distrust of intensive farms: “The grass the outside cows eat is different from the local cows; the milk they produce has a grassy smell, and the dairy products made from it have a strange taste.”
In their eyes, the shift in raw milk used by standardised workshops has a disruptive impact on the quality of the final dairy products.
IV. From Fresh Milk to Dairy Products: A Story from the Past
Such a possibility did once exist. In fact, from 2001, Zhengblue Banner, where Sangengdalai town is located, implemented an ecological migration policy paired with a dairy cow village construction plan. This guided migrating herders to abandon their local Mongolian Yellow cattle and transition entire villages to Holstein dairy cow farming.
Although imported Holstein cows were far more expensive than beef cattle and more difficult to care for, the government initially provided active loan subsidies, established fodder bases, introduced large dairy enterprises, and set up milk stations in every gacha. For a time, the local dairy farming ecosystem flourished, and the livelihoods of the herders became inextricably linked to it.
Take the Taben Aodu gacha in Sangengdalai as an example: around 2007, the average household owned 10 to 20 dairy cows, and the purchase price for milk remained stable at 1–1.2 yuan per jin. It is said that raising dairy cows could bring in dozens of yuan a day for a household—roughly equivalent to the daily wage of a casual labourer.
However, after the 2008 melamine “poisoned milk powder” scandal, regulations on raw milk quality tightened abruptly, and the dairy market shrank rapidly. Heavily impacted large dairy companies intervened directly in the raw milk market, establishing their own corporate farms and seizing bargaining power by absorbing individual herders. As the milk stations disappeared, the milk lost its original sales channels, and herders lost a vital source of income.
At the same time, the ripple effects of the 1980s grass-and-livestock contracting system on herders’ livelihoods began to emerge.
As pastures and livestock were allocated to individual households, the original production method of the grassland regions transitioned from collective “nomadism” to “managed pastoralism”, with households grazing within designated plots and bearing their own profits and losses. Since 1990, despite the implementation of ecological protection policies such as returning farmland to grassland, grazing bans, fallow periods, and grass-livestock balancing, the problem of pasture degradation remained prominent.
When the majority of pastures in Sangengdalai entered these grazing ban and fallow periods, the average usable pasture per person dropped to just over two hundred mu. This made it difficult to meet the conditions for free-roaming livestock, leading to a gradual shift toward fenced confinement, with herds being penned for over half the year on average. This meant a significant increase in both labour input and raw material costs.
For herders who move to cities for work, the cramped living environments are unsuitable for raising livestock, and even less so for making traditional dairy products.
Those herders who remain in the grazing areas and dairy villages are ageing; they milk their cows daily only for their own consumption and rarely sell milk to dairy processing workshops. They have also gradually abandoned the arduous labour of traditional dairy production, and their need for commercial dairy products has become increasingly urgent. This feels somewhat ironic.
V. Epilogue
The introduction of official production standards has indeed improved the hygienic conditions of dairy production, but increasingly standardised regulations have come into conflict with traditional knowledge of dairy making. Operators are forced to modify “old methods” to comply with production standards—but how can they then produce the authentic and delicious dairy products that herders hold dear?
Over time, commercial dairy products squeeze out traditional ones, big-brand dairy companies squeeze out individual producers, and the divide between those with means and those without becomes harder to bridge. Amidst the volatility of the raw milk market and corporate manipulation, the “dairy industry revival” is proclaimed with fanfare, yet small producers find it difficult to share in the prosperity.
To this end, Grile used a Douyin video to issue a call, requesting that regulatory departments introduce policies and measures better suited to traditional dairy workshops.
I believe that, beyond moving toward scaling and standardisation, respecting indigenous local knowledge and selectively collaborating with the natural environment and microbes is one possible path for traditional dairy products to emerge from their current predicament.
Foodthink presents the 2024 Lianhe Creation Project
Funding 5 to 10 creative works focused on food and agriculture
Encouraging content creators and researchers to explore the complexities of food and agriculture
Bringing more attention to the stories behind the journey from land to table
▼
Disclaimer
All names appearing in this article are pseudonyms
Unless otherwise stated, all photographs were taken by the author
Editor: Ze En