Let’s think about (left) liberalism over the last 50 years. By the 1960s, income had become relatively equally distributed. Liberal economists grew increasingly skeptical of labor unions. When the UAW demanded higher wages, the money seem to come out of consumer pockets, not corporate profits. By the 1990s there were increasing doubts about the welfare state. The war on poverty was widely seen to have failed. Bill Clinton said we needed to end welfare as we know it. Also in the 1990s, fiscal policy went out of style. New Keynesians believed the Fed had the tools to steer a fairly steady growth path in aggregate demand using Taylor rule-type techniques. Fiscal stimulus gets offset with tighter money in that sort of policy regime, leaving nominal GDP unchanged. Fiscal stimulus conceivably might have some supply-side channels, but liberals never put much weight on those effects.

I don’t think liberals were happy about this state of affairs. They tend to be optimists who believe government spending can fix societal problems. The neoliberalism of the 1990s represented a sort of rejection of that optimism. In the last 10 years, however, old-style liberalism has made a big comeback in two areas; fiscal stimulus and income redistribution. Here I’d like to see if there are some hidden connections between those two areas.

Two real world factors have led to a revival of liberalism: near-zero interest rates and increasing income inequality. When rates fell close to zero in Japan, and then later in the US and much of Europe, there was a revival of interest in fiscal stimulus. In retrospect, I think that was a mistake. According to Paul Krugman the year 2013 provided a test of the hypothesis that austerity slows economic growth. The outcome of the test couldn’t have been more clear-cut; American GDP growth actually accelerated in 2013. Even worse, European austerity occurred during a period where the eurozone wasn’t even at the zero bound. However as Mark Sadowski recently showed, that didn’t stop well respected pundits such as Simon Wren-Lewis from claiming that austerity had clearly slowed growth. And that was despite the fact that there is no respectable Keynesian model were austerity significantly slows growth during a time period where the central bank is raising interest rates.

The renewed interest in income redistribution also has its problems, but at least has some firm theoretical underpinnings. There are respectable utilitarian models where increasing consumption inequality calls for increased redistribution. Unfortunately liberals went way overboard, calling for higher minimum wages, higher taxes on capital income, and other extremely dubious remedies.

Here’s a common theme I see. Most liberals prefer to think like accountants, not economists. The dismal science focuses too much on the “no free lunch” concept. The idea that there are trade-offs, that incentives affect behavior. The idea that making failure less costly, also makes it more likely to occur. But liberal economists are not stupid, and as the 1990s demonstrate they are willing to adjust their policy prescriptions to reflect changing information about the market system.

Many liberals honestly believe that the past 10 years demonstrates that there are two giant pots of money just sitting there, of which they were not aware (or which were not available) in the 1990s. These pots of money can be spent on all sorts of socially worthy projects at almost no cost. One pot of money is fiscal stimulus. In the multiplier model there’s no opportunity cost of using fiscal stimulus, even on wasteful projects like high-speed rail between Tampa and Orlando. Indeed the multiplier effect creates a sort of “opportunity benefit” in the form of extra income above and beyond the government spending. That’s a free lunch. Liberals love free lunches. Actually, I love free lunches, I just don’t see as many out there as they do.

Fiscal stimulus is not the only free lunch. The increase in income inequality combined with a perception that many high earners are actually collecting unearned rents combined with studies suggesting minimum wage increases may not increase unemployment, have created another huge pot of money to play with. A free dinner. Liberals tend to assume that beyond some point the marginal utility of consumption is very low, especially for the ultra-wealthy. I agree with this. I’m much more skeptical of the minimum wage argument, and also the view that high CEO incomes are mostly unrelated to expected productivity. Both arguments seem highly implausible.

In 1986 Ted Kennedy helped enact a top income tax rate of 28%, despite opposition form a number of GOP senators. In 1987 the New York Times advocated abolishing the minimum wage. In the 1990s lots of liberals favored replacing welfare with workfare. Lots of liberals favored replacing taxes on capital income with a progressive consumption tax. They opposed fiscal stimulus. They thought unemployment compensation created unemployment—it made people lazier. These were all good ideas, and still are. I don’t know whether to be bemused or annoyed by the fact that today’s liberals think any right-winger that holds these views (that the best and brightest liberals once held) are stupid. Just because they’ve “jumped the shark” doesn’t mean that all the rest of us must follow.

In the 1980s, supply-siders took a good idea and went way too far in downgrading the importance of demand-side issues. Today’s liberals are taking a good idea (increased inequality might require more redistribution) and imagining this opens the door for endless free lunches and dinners. The supply-side doesn’t matter. If we only put higher taxes on the rich there would be a big pot of money for single-payer healthcare. A week later they forget they’ve spent the money (in their minds) and now want to spend it on free preschool education. Another week goes by in that same pot of money can be spent on improved infrastructure. The same money can be spent 100 times. The “unmet needs” are endless.

I’m only about 250 pages into the Piketty book, but it already seems to epitomize modern (American) liberalism. So far it seems 99% accounting and 1% economics. In the few times where incentive effects come up they are generally dismissed as being unimportant. For instance we are told that Thatcher’s economic reforms didn’t really turn around the British economy (hard to accept if you’re old enough to recall how poorly Britain was doing in 1979.) It’s the kind of argument that appeals to people who just look at times series of average GDP growth rates, and not how Britain was doing relative to France in 1979.

Paul Krugman likes to say that reality has a liberal bias. But economics has a sort of conservative bias, or more precisely a free-market bias. It’s innately hostile to arguments that there are big pots of money that can be used without opportunity cost. That doesn’t mean liberal arguments are always wrong, I do think there are some good arguments for redistribution. Rather I would argue that there are certain times when events get misinterpreted and people overestimate how much can be accomplished with liberal remedies. The 1930s was one of those times and the period since 2008 is another. Big government liberalism is an ideology built 20% on reality and 80% on cognitive illusions.