Democratic Planning – Part 4

Introduction:

(For previous parts in the series, see here.)

For all that we think we live in a free market, the truth is that we plan all the time. As Matt Yglesias and others have pointed out, zoning is planning, and so is licensing of certain kinds of businesses (bars, restaurants, etc.) and all the other kinds of regulations that are so ground-level they become almost invisible.

The problem is that the kind of planning we do is fragmented, not systematic, unconscious not conscious (or at least not openly conscious – consider how bans on multiple-occupancy dwellings in suburbs are sold as “protecting housing values” as opposed to “keeping out poor people”, and not particularly intelligent.

And this won’t change until we start to learn from our actions.

Cash for Clunkers as Model:

Because most economic planning done in the United States is so fragmented, unconscious, and unthinking, it’s hard to find good case studies to learn from. Luckily, with Cash for Clunkers, we have the rare example of a very discrete experiment with economic planning that can be more or less evaluated on its own. Because in one aspect, the wingnuts are right – Cash for Clunkers is an attempt to plan for the expansion of the auto industry, and the transition from a low-mileage industry to a higher-mileage industry – and it’s actually a good thing.

As I’ve noted before, Cash for Clunkers was rather unfairly maligned. As an experiment, the results were fairly promising. According to the U.S Treasury, the program, which only ran for two months, spent $3 billion in subsidies, resulted in the sale of 700,000 cars, produced 42,000 jobs, and created .3-4% worth of GDP. The public purpose of the program was rather successful – the average fuel economy of the cars traded in was 15.8 mpg, whereas the average fuel economy of the new cars bought was 25.4 mpg.

As an example of Keynesian planning, how does Cash for Clunkers hold up?

On the positive side, the program has a pretty good bottom line – $3 billion for .3-4% of GDP and 42,000 jobs is good for stimulus, working out to a multiplier of about 12 for GDP and about $70,000 per job created. The cost-per-job is a bit high (it would cost $72 billion to create 1 million jobs, versus about $35 for a jobs program), but in terms of GDP creation, a Paul Krugman or Brad DeLong would say that it’s a pretty good rate. And as I have noted, a strong public purpose (in this case, improving fuel efficiency) is essential for economic planning’s legitimacy.

However, Cash for Clunkers also had important limitations that have to be understood and confronted. To begin with, the program’s termination shows a basic limitation of both scale and scope. Only funding the program at $2 billion, exhausting in two months what was supposed to last for five, shows that the administration really mistook how much demand there would be for the program. Moreover, there is a basic problem with planning for a five-month boost in sales at a time when nigh-double-digit unemployment and the after-effects of the credit crunch have led people to dramatically dial back spending -without a government intervention that actually plans past the positive impact of new sales and the need to restock cleared lots, or that envisions a wide enough recovery of consumer spending to take up the slack after the program’s expiration, you’re likely to see a “dead cat bounce.”

Moreover, while the program’s credits (and the fact that it was tied in to trade-ins) reduced the cost of buying a new car to consumers, Cash for Clunkers didn’t create as much income as was needed, and in part relied on people spending their own money to make up the difference. While on one level that’s good for stimulus (getting people to increase their “propensity to consume,” decreasing their liquidity preference), it’s also the case that it could have the effect of shifting, rather than outright boosting income. That’s one of the reason why programs like Unemployment Insurance or Food Stamps (classic “automatic stabilizers”) are considered top-quality stimulus levers, because the income they create is 100% new, rather than requiring levering existing income.

This all points to a large lack of vision with the program. Cash for Clunkers was clearly designed as an emergency program to protect the newly-rescued automotive industry, not for the long-term. It wasn’t done with any view in mind to boost income levels and consumption over the long term, and it certainly wasn’t done with a view in mind as to what the future size of the automobile industry should be.

Overall grade? Call it a B/B+.

Lessons:

So what larger lessons can we learn from Cash for Clunkers?

• Planning Has to Be Done At the Right Time Scale:

It doesn’t do much good to do a whole bunch of stimulus at the beginning of a recession, only to cut off the priming-pump right when things turn around, causing a “double-dip” recession. Similarly, it doesn’t do much to boost the production of a single industry for five months unless you can be sure that there’s going to be enough demand on the other side – otherwise, the effort put into boosting production levels, opening up new plants, and rehiring workers will go to waste when demand falls back down and plants close back up.

To that end, planning has to plan from the trough of a recession well into the recovery, beyond the immediate business cycle.

To once again use the Cash for Clunkers program, you need to plan out the point where demand for cars will have recovered (which at the very least will be until unemployment levels and thus consumption levels return to “normal”).

• Planning Needs to Mobilize Unused Resources:

One of the reasons why Keynes emphasized that stimulus should be financed by borrowing and/or progressive taxation and then spent on government programs that help out poor people is that one f the economic maladjustments that stimulus is supposed to cure is a mal-distribution of income between consumption and investment and between the poor (who have a high “propensity to consume”) and the rich (who have a low propensity), which is more or less the same thing, since most consumers are not rich and most investors are. The effect is to move capital from its unproductive state (either being held liquid by people whose liquidity preference has zoomed through the ceiling because of recession-induced anxiety, or being overly-concentrated in the financial sector) to boost consumption and create a source of profits for productive investments.

To that end, planning should seek to draw resources from where they are not being used (in this case, including the financial sector and higher-income groups as well as consumers); attempting balanced budgets in the middle of a recession, for example, would be really bad planning. A loan from the Fed would also have done the job rather nicely.

• Planning Should Be Holistic In Its Approach to the Economy:

Arguably, one of the major shortcomings of the last twenty years worth of economic policymaking has been a blinkered approach to industries and their relation to the overall economy. The economic “recovery” of the mid-2000s, for example, was heavily concentrated on real estate, financial services, and debt-fueled consumer spending in part because that’s where the government’s policy of low interest rates and financial deregulation pushed it. Similarly, in the late 1990s, the high-tech and e-commerce sectors were largely relied on for most of the economic and job growth of that era.

However, planning should keep in mind how all of these different parts fit together. In that sense, one of the objectives of economic planning should be to balance existing industries to keep one sector from becoming too large and distorting the economy – such as an overly-large financial sector destabilizing the real estate or energy sectors. Similarly, economic planning should also seek to create new growth and jobs, not by relying on a single industry – incidentally, that’s one of the reasons why the current progressive fixation on “green jobs” as a panacea is rather dangerous – but in seeking expansion in a number of industrial sectors, to ensure that economic progress is widespread, and the benefits spread throughout the economy.

Conclusion:

In the end, policy is not as magical as policy experts would like people to believe. You can learn from doing it right and from doing it wrong, and ultimately I believe that FDR was onto something when he proclaimed that “bold, persistent experimentation” would be the hallmark of his administration. After all, how often in the last half-century has an overly-rigid dedication to one particular theory lead us astray?

However, what is required in order for this to work is a genuine commitment to transparency (so we can see what we’re doing) and self-awareness (so that we’re aware of what we’re doing) and a genuinely open political discourse (so that we’re willing to entertain a diverse plurality of views as to what works, what doesn’t, and why).