In a justly criticised piece in the FT, Morgan Stanley's Ruchir Sharma says:

My team also found a statistically significant link between periods of rising government debt and slow GDP growth. These studies cannot show causation, but the consistent link between growing deficits and weakening growth is unlikely to be coincidence.

True, it's unlikely to be coincidence. What is likely, though, is that the causality runs from a weak economy to rising government debt. Slow growth causes rising government debt not just because it depresses tax revenues and raises welfare payments but also because it causes governments to loosen policy.

Sharma's failure to give due weight to this obvious possibility is an egregious example of a widespread error - an inability to appreciate that policy is endogenous. This arises from the very language of economics. We tend to speak of policy as "interventions", as if policy-makers were outside the economic environment, stepping in only to meddle or to correct a market failure (delete according to ideology). But in fact, policy-makers are an inescapable part of the environment, and their beliefs and actions are influenced by it and vice versa.

Another example of this type of error is when it's said - often by a type of Austrian economist - that QE and zero interest rates have pushed asset prices up to "artificial" levels, as if policy were an unnatural deviation. But this is wrong on two counts. For one thing, the only prices that matter are those at which assets actually change hands; it is your imaginary counterfactual world that is artificial. And for another, it's likely that policy measures to reduce interest rates only reflect economic forces such as low growth and capitalists' reluctance to invest.

That error reflects a wider and longer-standing one - one that dates back to Smith's claim that humans have a natural "disposition to truck, barter, and exchange". It's the idea that markets are natural and thus that interventions are unnatural. But this is not the case. Markets were, wrote (pdf) Polanyi, "incidental to economic life" before capitalism, and a market society is, wrote David Graeber, a "relative newcomer.": Ellen Meiksins Wood is also good on this. There's nothing artificial about state involvement in the economy; it's been going on for centuries.

It's not just people with daft ahistorical ideas about markets who fail to see the endogeneity of policy. Here are some other examples.

1. A few years ago, the Independent ran a series in which people said what they'd do if they were PM. This told us plenty about the writers, and nothing about politics. The Prime Minister and his policies don't fly in from outside to rescue us like the Thunderbirds. They are the outcome of socio-economic structures that select for policies and individuals, and not necessarily the best.

2. Sarah O'Connor describes how "some benefits of a higher minimum wage can leech away in an under-regulated labour market" as employers cut overtime pay or hours or simply fail to comply with it. What this omits is that such under-regulation is no accident. It's endogenous - a consequence of the power of capital to resist what few demands there are for tighter labour market regulation.

3. Several recent books, such as Lindsey and Teles' The Captured Economy, Case and Deaton's Deaths of Despair and Thomas Philippon's The Great Reversal describe the rise of rent-seeking and cronyism in the US and advocate policies to counteract it. It is, however, wrong to think that rolling back cronyism is merely a matter of policy-makers knowing what to do. It also requires a favourable balance of class power. And right now, (some) capitalists have the power (pdf) to bend government to their will. Crony capitalism is endogenous - the result of capitalist power. Reversing it significantly requires a diminution in that power.

4. Donald Trump did not become president because Americans suffered a collective knock on the head from which they recovered last November. His popularity is instead the result of socio-economic conditions. The "white anxiety" which he exploited is, says Jonathan Metzl, "a problem, not just of individual minds or attitudes, but of larger social and socioeconomic structures." Part of the story here is that these structures contributed to industrial decline and a loss of income and status among some voters. Trumpism, says Samuel Farber, is "a right-wing response to the objective conditions of economic decay." And, adds Branko Marcetic, it is "the logical, some would say predictable, product of the failures of not just Barack Obama, but decades of neoliberal politics." Back in 2006 Ben Friedman showed that economic stagnation fuelled racism, intolerance and anti-democratic sentiment. Trumpism vindicates that view. Unless these socio-economic conditions are significantly changed, we risk seeing a re-emerging of some form of Trumpism.

My point here is to challenge a particular view of politics which is perhaps widespread among centrists. It's the belief that we could have better policies if only we were governed by people of intellect and goodwill. But this misses a big fact - that actually-existing capitalism tends to militate against good government by, for example, promoting cronyism, managerialism and reaction and selecting against technocrats. Liberal, technocratic government requires that we challenge capitalism.