The Grip and the Sand

How the Fear of Losing Control Becomes a System for Losing What Matters

There is an old physical stupidity that almost everyone understands instinctively. Hold dry sand loosely in your palm and most of it remains there; become frightened that you are losing it, close your fingers, then tighten them again, and the grains begin escaping through every available gap. The natural response is to squeeze harder, because the person holding the sand interprets every escaping grain as proof that the previous grip was insufficient. Before long, the hand is no longer preserving the sand. It is producing the loss it fears.

This is a useful way to read the direction of Chinese governance in recent years, because what looks from the outside like an assortment of unrelated policies—capital controls, regulation of cross-border data, technology-export restrictions, tighter supervision of overseas investment, cybersecurity rules, political demands on private enterprise—can also be understood as expressions of the same institutional anxiety. Beijing sees more things moving beyond its immediate reach, interprets that movement increasingly through the language of security, and responds by extending the perimeter of political control. The immediate result may indeed be more control. The longer-term result can be a society with fewer reasons to create the things worth controlling.

China is hardly unique in regulating capital, technology or data, and pretending otherwise would weaken the argument. The revealing feature is not the existence of any particular regulation but the expanding tendency to connect economic activity to political security. The revised Cybersecurity Law explicitly places cybersecurity under Communist Party leadership and the “overall national security concept”; China continues to regulate cross-border transfers of data through security assessments and other compliance mechanisms; and the 2023 technology-export catalogue still contains 24 prohibited and 110 restricted categories after revision. None of these facts alone demonstrates economic decline. Together, however, they illuminate a political system whose definition of what must remain governable has steadily expanded.

The paradox becomes particularly visible in the treatment of private enterprise. Beijing knows that it needs entrepreneurs, investment and technological experimentation badly enough that in 2025 it enacted the country’s first dedicated Private Sector Promotion Law. The law promises fairer market competition, financing support, innovation protection and safeguards for private businesses. Yet its second article establishes that promoting the private economy must uphold Communist Party leadership and maintain the “correct political direction,” while another provision specifies a political role for Party organizations operating inside private firms. The contradiction is almost too elegant: the state wants the economic consequences of autonomy while remaining suspicious of autonomy itself.

This is where the handful of sand matters, because the central mistake of an insecure authoritarian system is to confuse retention with vitality.

Capital that has difficulty leaving has not necessarily become productive capital. An entrepreneur who cannot easily move abroad has not necessarily become an ambitious entrepreneur at home. Technology prevented from crossing a border does not automatically reproduce itself. Information successfully contained does not become knowledge. A company that remains registered inside the jurisdiction can still stop taking serious risks, while an investor whose money remains inside the country can choose safety, liquidity or political proximity rather than finance another uncertain experiment.

A bureaucracy can therefore become remarkably successful at keeping the visible pieces of an economy in place while gradually damaging the invisible incentives that made those pieces valuable.

The distinction is crucial because the state can command location much more easily than motivation. It can determine where money may move, what data may cross a border and which technologies require permission to leave, but it cannot order someone to remain curious for another decade. It cannot regulate an investor into confidence, instruct a scientist to become intellectually reckless, or command an entrepreneur to wager years of life on a strange idea whose commercial value nobody yet understands. The most productive forms of economic behavior depend precisely on a zone of uncertainty in which nobody, including the government, knows beforehand what will work.

For a political system organized around control, that uncertainty eventually becomes uncomfortable. A successful private company is economically useful but politically less predictable than a state organization. International capital is useful but mobile. Globally connected researchers are valuable but difficult to isolate. The internet is commercially productive because information moves cheaply, yet that same property makes information harder to contain. Overseas expansion gives companies access to markets and knowledge, but also gives companies alternatives. Every mechanism that increases economic adaptability therefore creates some corresponding reduction in political legibility.

During periods of confidence, an authoritarian government may tolerate this bargain because growth makes the disorder worthwhile. During periods of insecurity, however, the same ambiguity begins to look different. Mobility starts looking like flight; independence like disobedience; international connection like foreign influence; unsupervised information like vulnerability; corporate autonomy like political risk. The government does not need to announce that economic freedom has become undesirable. It merely has to expand the number of circumstances in which freedom requires permission.

The feedback loop then becomes self-reinforcing:

insecurity → tighter control → lower dynamism → greater insecurity → tighter control.

The important part of this sequence is not the first tightening but the second. A government becomes worried about capital leaving, so it restricts movement; businesses then attach a larger political-risk premium to future investment, which weakens investment and gives the government more reason to worry about economic stability. Officials become concerned about technological leakage, so technology becomes more tightly governed; researchers and companies respond by avoiding ambiguous projects or international arrangements, which reduces experimentation and intensifies anxiety about technological competitiveness. Authorities become concerned about information, so information becomes more controlled; the resulting deterioration in reliable feedback makes the state itself less capable of distinguishing genuine danger from bureaucratically manufactured reassurance.

Control creates the symptoms that justify more control.

This also explains why the process does not require stupidity at the level of individual officials. Quite the opposite: every individual decision can be rational inside the political system that produces it. A security official is rewarded for eliminating a security risk, not for estimating how much entrepreneurial confidence will disappear five years later. A regulator can measure whether capital left the country more easily than whether an investment was never attempted. A censor can count information removed but cannot count ideas that were never conceived because the intellectual environment became narrower. The benefits of control are immediate, visible and attributable; the costs are delayed, dispersed and often counterfactual.

The institutional bias therefore points in one direction even when nobody explicitly chooses the final destination.

This is also why attempts to repair the economic consequences can appear strangely contradictory. The state can sincerely want more private investment while simultaneously strengthening political supervision of private enterprise. It can sincerely want world-class technology while making technological exchange more politically sensitive. It can sincerely want companies to expand internationally while becoming more concerned about what capital, knowledge and people carry across borders. There is no need to assume that one side of this contradiction is fake. Both desires can be genuine, because they belong to different imperatives inside the same system: the economy requires movement, while the political structure increasingly demands governability.

The problem is that only one of those imperatives possesses coercive power.

When economic dynamism conflicts with political security, the entrepreneur cannot order the security apparatus to accept more uncertainty. The investor cannot require the Party to tolerate greater institutional autonomy. The researcher cannot demand that the state narrow its definition of strategic risk. Political control therefore tends to win each individual dispute even when the accumulation of those victories makes the economy less capable of producing what the state wants.

That is the deeper danger of the tightening fist. It is tempting to imagine authoritarian failure as a spectacular moment in which the government suddenly becomes too weak to control society, but a state can also damage itself through the opposite process: it can become extraordinarily capable of controlling the measurable surface of society while progressively reducing the spontaneous activity underneath it.

Nothing dramatic needs to happen. Factories can continue operating, banks can remain open, companies can remain registered, engineers can continue going to work and the government can retain formidable administrative capacity. Yet investment horizons shorten, politically safe sectors attract more resources, entrepreneurs become more cautious, researchers learn which questions create trouble, wealthy families diversify whatever they still can, and talented people devote increasing intelligence to navigating rules rather than discovering possibilities. The system remains standing while becoming less generative.

That distinction matters because predictions of imminent collapse are both unnecessary and analytically distracting. The tightening of control does not tell us when China will experience a political rupture, and history offers plenty of examples of rigid systems surviving far longer than outsiders expected. The more interesting question is not how long the fist can remain closed, but what happens to the contents while it does.

The Communist Party’s fear is understandable on its own terms. Capital can leave. Information can undermine authority. Technology can migrate. Companies can become independent centers of influence. People with alternatives are harder to govern than people without them. From the standpoint of a Leninist political organization, each of these developments contains a genuine political risk.

But the things that make a modern society powerful are often inseparable from precisely those risks. Capital is useful partly because it can choose. Knowledge advances partly because it crosses boundaries. Entrepreneurs create value partly because they make decisions bureaucracies would not make. International networks are productive partly because no government completely controls them. Innovation matters precisely because its destination cannot be specified in advance.

A government can reduce these uncertainties, but it cannot remove only their political consequences while preserving all their economic benefits.

Eventually the choice becomes visible: tolerate movement and accept that some things will escape political control, or eliminate more movement and discover that control has preserved the container while diminishing what made the container valuable.

The tragedy of the handful of sand is therefore not that the hand is powerless. It is that the hand is powerful enough to keep tightening. Every grain slipping through the fingers confirms the fear that produced the previous squeeze, and every squeeze creates another reason to fear what will happen if the hand ever relaxes.

Beijing does not need to lose control for this cycle to become dangerous.

It only needs to become increasingly afraid of losing control.