The following is a slightly edited version of a set of articles that appeared on NewDeal20.org, the blog site for the Eleanor and Franklin Roosevelt Institute.

Paul Krugman recently argued that “manufacturing is one of the bright spots of a generally disappointing recovery, and there are signs — preliminary, but hopeful, nonetheless — that a sustained comeback may be under way.” He points out that the gap between what we sell and what we buy has been improving. This must be set against a background of a manufacturing decline in the United States of historic dimensions; even without adjusting for inflation, the trade deficit in goods for the United States between 2000 and 2010 was 7 trillion dollars. A turnaround in the attention of more perceptive economists and a turnaround in manufacturing may be in the works. But before that, the crucial question is: Why is manufacturing so important?

1. Manufacturing has been the path to development

It has been the strategic achievement of rich nations over the last several hundred years to create a high-quality manufacturing sector in order to develop national wealth and power, as Erik Reinert shows in his book “How Rich Countries Got Rich…and Why Poor Countries Stay Poor.” From the rise of England in the 19th century, to the rise of the US, Germany, Japan and the USSR in the 20th, to the newly industrializing countries like Korea, Taiwan, and now China, manufacturing has been the key to prosperity

2. Manufacturing is the foundation of global “Great Power”

The most powerful nations in the world — the “Great Powers” — are those that control the bulk of the global production of manufacturing technology. That is, it isn’t enough simply to have factories and produce more goods, you have to know how to make the machinery that makes the goods. The key to power, then, is to make the “means of production.”

As the machinery industries go, so goes Great Power. My own research shows that about 80% of the world’s production of factory machinery has been controlled by what we would consider the “Great Powers.” Until the 1950s, the US had produced about 50%; we now produce less than China’s 16%.

3. Manufacturing is the most important cause of economic growth

The growth of manufacturing machinery output, and technological improvements in that machinery, are the main drivers of economic growth. No machinery industries, no sustained, long-term economic growth. Just consider the explosion of the Internet, iPhones, and the like — all made possible by a small subset of production machinery called semiconductor-making equipment (SME), which itself is dependent on other forms of production machinery, such as the machine tools that grind the lenses they use or the alloys of metal the metal-making industries output. These technologies reproduce themselves, as when an SME makes the semiconductors that then go to make more SMEs, or when a machine tool makes the metal components that not only go into other pieces of machinery, such as cars, but are used to produce yet more machine tools. The technological and productive potential of machine tools and SMEs affect each other as well, leading to the explosive economic growth of the last two hundred years.

4. Global trade is based on goods, not services

A country can’t trade services for most of its goods. According to the WTO, 80% of world trade among regions is merchandise trade — that is, only 20% of world trade is in services. This closely matches the trade percentages that even the US, allegedly becoming “post-industrial,” achieves. If in the extreme case an economy was composed only of services, then it would be very poor, because it couldn’t trade for goods; its currency would be worth very little. The dollar is also vulnerable in the long-term. A “post-industrial” economy is really a pre-industrial economy — that is, poor.

5. Services are dependent on manufactured goods

Services are mostly the act of using manufactured goods. You can’t export the experience of using something. Retail and wholesale, which make up about 11% of the economy, are the act of buying and selling manufactured goods. The same goes for real estate, another 13%, which is the act of buying and selling a “real” or physical asset, a building. Even health, which makes up about 8% of the economy, is the act of using medical equipment and drugs (all figures from 2010, value-added).

Finance involves the redirection of surplus resources that the nonfinancial sector of the economy produces, which means that indirectly, even finance is dependent on manufacturing. The cycle of rise and decline usually runs like this: some clever society figures out how to take advantage of the current technologies of production, thus generating huge surpluses, which either the financial forces, the very wealthy, or the military then appropriate for their own wealth and power; they kill the goose that is laying the golden eggs. To sum up: the health of the economy is critically dependent on the health of the manufacturing sector.

6. Manufacturing creates jobs

Most jobs, directly or indirectly, depend on manufacturing — and reviving the sector could provide tens of millions of new jobs, eradicating the Great Recession. In 2005, the Japanese manufacturing sector was 20.2% of its economy, in Germany it was 23.2%, and in the US manufacturing accounted for 13.4%, according to the the OECD. Using 2005 figures, if the US had the same percentage as Japan, we would have 7 million more high-quality, long-term, well paying jobs. If we were equal with Germany, we would have 10 million more. And according to the Economic Policy Institute, each manufacturing job supports almost three other jobs in the economy. That makes sense, considering the other five reasons that manufacturing is central to the economy.

Thus, there are six solid reasons that we need to rebuild the manufacturing sector of the United States. Unfortunately, the reigning economic theory minimizes the importance of manufacturing.

The Economy is an Ecosystem

Question: Why are we using mainstream economic ideas rooted in the 19th century to deal with 21st century challenges to our global civilization and to our national economies? No wonder political debates can seem like running in mud. Progressives are constantly being dragged down into a fantasy world of perfect markets, balanced budgets, and very limited government intervention. Neoclassical economics was not designed to deal with the complexities of a global industrial civilization that is endangering its own existence ecologically with global warming and ecosystem destruction, and economically with an out-of-control financial sector. It wasn’t designed to solve the problems of the world central economy, the United States, and its seriously declining manufacturing sector.

Any efforts to extricate ourselves from outdated economic thinking gets the basic response: whatever needs to be done can’t and shouldn’t be done by the government. Should the government support a revival of manufacturing which is central to the economy? No, is the answer. The market will work itself out. And if the market has decided that manufacturing should be exported to countries outside the US, then the market must be right. Should we prepare for a society in which oil is much, much more expensive? No, comes the response. The market will take care of that, too. If oil becomes too expensive, something else will magically pop up to replace it. Should the government create new jobs since we have a huge unemployment problem? No, the market knows when to hire. In fact, the thinking goes, since taxes can only prevent the market from setting the best price, all taxes are suspect, and all regulation is, too, even for a financial sector that caused a global depression.

The only way out of this intellectual straightjacket is to construct an entirely new paradigm of how the economy works. Here’s a start: what if the economy is really an ecosystem centered , say, on manufacturing? If we looked at it this way, we would think very differently about the way governments intervene in its functioning.

An economy has many parts – a financial system, a manufacturing sector, a transportation system, natural ecosystems, and many others. Each part is, in turn, made of many other parts. In the neoclassical world, the economy is divided up into individual firms – and these are not differentiated in any meaningful way. In every introductory economics class, the professor explains that many identical firms exist in a perfectly competitive industry, and that the interaction of these firms leads to a particular price for a product.

As Philip Mirowski showed in his classic work, “More Heat than Light”, the men who created neoclassical economics were very much influenced by the “hot” science of the day, statistical mechanics. If engineered properly, a gas or fluid system will remain stable. Such a system (for example, your plumbing) will settle back down to a stable state even after being disturbed. Statistical mechanics achieves its great power because it is able to treat each part of a system as identical. There are no categories or large, differentiated parts of such a system; all water molecules, for example, are the same.

What has this got to do with conservative economic activities, such as corporate lobbying against regulations? The lobbyists use the argument that, left alone, like a plumbing system, a market will settle down to “the best of all possible worlds” – left alone by the government, that is. Many economists began academic life in the physical sciences – Mirowski relates how Leon Walras, the creator of equilibrium analysis, the bedrock of current economics, was an engineer who kept a copy of a statistical mechanics textbook next to his bed. If many other branches of academia have “physics envy” – that is, they can’t predict with the precision and elegance of physics – then economists have grabbed the golden ring, and try to claim “scientific” accuracy.

Except, of course they aren’t accurate. As Dean Baker has pointed out numerous times, the vast majority of economists missed the recent housing bubble. Economies are not made up of identical parts – they are made up of a plethora of very different parts, and like ecosystems, these parts can be thought of as niches, that is, as functionally differentiated parts of an ecosystem, all of which are necessary for the growth and health of the whole. Even if a part of an ecosystem – say, manufacturing, in the case of an economy – does not comprise the majority of the production of the ecosystem, that does not mean that the part is not the most important, or the source of the most important changes. Even a very small niche – say, machine tools in an economy – can have a very oversized effect on the system as a whole.

In the 20th century biological systems, such as ecosystems, increasingly became the focus of scientific research. Even in the field of mechanics, scientists such as Ilya Prigone helped to invent the field of “chaos theory”, or nonlinear dynamics, that is, the study of systems that are not stable. Evolutionary theory, first developed by Charles Darwin, showed that you could have a science that did not predict exact outcomes –- how can you predict the creation of a new species, or ecosystem? The interaction of discrete niches and species, all different, does not lead to complete “chaos”, it is possible to make some broad predictions about how an ecosystem will develop, thrive, and die.

In the same way, we can make statements about how an economy will develop, thrive, and die, when we view an economy as a set of identifiable “chunks”, like manufacturing, or transportation systems, or natural ecosystems like rivers and the climate, and seek to understand how these parts fit together. When the way that a set of parts fits together effects how the set of parts act together, then we know that we are looking at a system — that is, as the saying goes, the whole is greater than (or different than) the sum of the parts (in the language of systems, we say that structure allows for new properties to emerge).

If each part of the economy –- including the natural ecosystems –- is important for the functioning of the entire economy, and if some ways of putting those pieces together are better than others, than it is the government’s responsibility to make sure that those pieces are thriving and fitting together in a productive way. The market cannot design the system as a whole. If, on the other hand, the pieces of an economy are interchangeable – if manufacturing is just as valuable for the economy as, say, tourism – then the government has very little to do, as the neoclassical viewpoint would have it.

Manufacturing is the foundation of economic growth

Underlying most debates about economic policy lurks a single question: what causes economic growth? This question is key when an economy is stagnating or declining. The answer decides the fates of Presidents, political parties, and whole nations.

Surprisingly, neoclassical economists have not had much luck explaining economic growth. Nobel-laureate Robert Solow, who is the most important growth theoretician, calculated in the 1950s that neoclassical economics could only explain about 20% of economic growth; the rest was “technological change,” which has been impervious to neoclassical analysis ever since. And yet, according to the economic historian Angus Maddison, between 1913 and 1989 income per person in most developed countries increased almost five times (see chapter 7 of my book, “Manufacturing Green Prosperity,” and also here). So how can neoclassical economics call itself a science if it can’t explain its single most important phenomenon in the economic universe? What if physics couldn’t explain why planets revolve around the sun?

The problem is that neoclassical economics relies on a model of reality that is similar to the model use by physicists to understand things like gases and fluids. There is no growth in a gas or fluid. When things change too fast in a gas, for instance, the system explodes –- it’s all very unstable and stressful. But neoclassical economics doesn’t have the analytical tools to explain growth because it bases its understanding of production on David Ricardo’s concept of “diminishing returns”. You can’t explain how something grows if you are using a concept that explains why something diminishes.

There is, however, a science that is very comfortable with the idea of accelerating growth – biology, and in particular, evolutionary biology and ecology. Organisms reproduce — bees do it, birds do it, and as I will argue, machinery does it, too. Reproduction leads to exponential growth, that is, steady growth that builds on itself, and that can even accelerate.

But how can we have growth forever and not destroy the planet — or, to be more precise, the biosphere? Again, ecology is comfortable with this issue, because it is a fundamental aspect of all ecosystems that there must be balanced growth. In other words, no one part of the ecosystem can run out-of-control and reproduce forever. No part of a sustainable ecosystem can produce something — like greenhouse gases or pollution — that destroys the ecosystem as a whole.

Well, then, how can we keep economic growth going and not destroy the ecological foundations of the economy? In both natural and economic ecosystems, there are two main sources of change: technological change, and change in the quantity of the output of the various parts of the system. In a natural ecosystem, technological change takes place in the form of “variations”, as Charles Darwin called them, of the offspring of organisms. We now know that DNA controls these changes. And depending on how these genetic “variations” fit into their environment, their populations may increase or decrease in size, which then feeds changes back into the ecosystem.

In a manufacturing ecosystem, the processes of change are similar. In order to understand this, we have to delve a little deeper into the “natural history” of manufacturing. The most important parts of the manufacturing system are the machinery niches, or industries. These are the ones that provide the equipment for the factories that output the manufactured goods that the society uses. When these technologies change, it becomes possible to create different kinds of goods, or to create more goods with the same resources. But how are these kinds of factory machinery made?

At the center of the economy resides a set of extremely important machinery industries –- let’s call them “reproduction machinery” –- which, if we continue with our ecological metaphor, can be viewed as collectively reproducing themselves. And since they can reproduce themselves, like birds and bees do, they can drive the economic growth of the economy, decade after decade, century after century.

Now you may be more familiar with the organisms that inhabit the volcanic vents in the deep Atlantic than you are with these kinds of machinery, but they are not that difficult to understand. There are machine tools, the machines that make parts (usually steel) that are used to produce most other machines, from cars to new machine tools. So machine tools can make more machine tools. However, they do it with help from other reproduction machinery – for instance, steel-making equipment, the huge pieces of machinery that take molten iron, carbon, and a few other elements and output the various kinds of steel that keep the current civilization running – such as the steel for the machine tools.

So we have machinery that creates the substance. We have material used for other machinery. And we have machinery that forms or shapes this material. Form and substance are two helpful categories of production that are nice to have control of if you want a modern civilization. Of course, we also need energy, and the main energy for making machinery is not oil, but electricity. We hope that machines like windmills can become the dominant way to make this electricity, but currently something called an electricity-generating steam turbine does the trick, formed out of steel, using machine tools. Once we have energy, we certainly want to be able to process information, and so we have semiconductor-making equipment, which makes the semiconductors that make all the other machinery better and better, including other semiconductor-making equipment.

When reproduction machinery gets better, everything else gets qualitatively better. When there is more reproduction machinery, there can be more of everything else – which becomes a problem if we are using everything up and destroying the environment and the climate. So we need to concentrate on making everything qualitatively better, not increasing everything quantitatively. This means living in moderately-sized living spaces, for example, and making them much more comfortable and efficient. It means using the same amount of space in a city to house more people, but making the city more comfortable and efficient. And it means using wind instead of coal, computers instead of paper, and electric vehicles that are more comfortable and efficient instead of oil-based ones. And recycling everything, just like a natural ecosystem.

The machinery industries are critical, both for economic growth in general, and in particular to create the kind of economic growth that is ecologically sustainable. They are more important for the optimal functioning of the economy than, say, the level of taxes or deficits or regulation. And yet, the United States has allowed its machinery industries to decline from 50% of world output to less than China’s 16%. Unfortunately, much of what remains is owned by foreign companies, and they do the high-value engineering outside the US. For the neoclassical economist, this is not a problem, because in the neoclassical world, the economy is global. But economies are actually based on world regions – like the United States.

Regionalization, not globalization

We often hear that a globalized economy is the best kind of economy. But could a world economy based on a set of strong regional systems, each one centered around a thriving manufacturing sector, serve us better in the long run?

Despite the rhetoric of globalization, the wealthiest economies have historically been regionally based. By “region,” I mean a geographically contiguous area, separated from others by a barrier. The premiere example of this is the United States. Europe is another natural region, which has now integrated itself formally, although it was always integrated in fact. Japan, China, and India are also always considered separate economies. In fact, the regions that have the least cohesion are the poorest, such as Africa. While Africa is a natural economic region, it has been “integrated” into the world economy at great expense because its various pieces have become resource-generating appendages of wealthy, regional economies, instead of remaining parts of a manufacturing-centered, integrated African economy. The same could be said for the Middle East; Latin America is somewhere in between and is part of the global “lower middle class” as a result. The post-Soviet set of countries of central Eurasia are poorer than their oppressive predecessor partly because they are not as integrated as they used to be.

What are the advantages of operating in a contiguous geographic area? After all, according to neoclassical economic theory, an exchange is an exchange is an exchange, whether it is between a customer and a local store or WalMart and a supplier in China. The problem is that an economy is composed of both exchange and production; you have to have something produced before you can exchange it. And like an ecosystem, production relies on many sub-networks of production and exchange that require a physically close set of interactions, particularly when it comes to innovation.

Innovation can be seen as the product of three main sets of “human capital” — scientists, engineers, and skilled production workers. Scientists create a “stock of knowledge,” to use Simon Kuznets’ phrase. This stock of knowledge is used by engineers, who design the machinery and processes that are used in the factories and construction sites and other sites of production. Skilled production workers then use these factories and other production centers to create the wealth that societies survive on.

Just as it would make no sense in a natural ecosystem such as a forest to have the trees in one place, the deer that eat the leaves in another place, and the bears that eat the deer in yet a third place, so it makes no sense to have the scientists, engineers, and skilled production workers spread out all over the globe, with little or no interaction. It is one of the great ironies of modern economic life that the industry that is perhaps doing the most to disperse these groups, the financial industry, is so geographically concentrated that it can simply be referred to by a single street, “Wall Street.” The financial industry concentrates itself for the same reason whole regions work best in a geographically bounded area: the main players can talk to each other, observe first-hand the processes that underlie their industry, and very quickly change practices as the larger environment, or ecosystem, puts pressure on the industry to change.

Engineering and research centers are now moving from the United States to China, just as factories have moved. These moves increase innovative capacity when researchers and engineers can interact with and witness first-hand the operation of machinery and factories, talk to other engineers and skilled production workers, and take advantage of the serendipity and unexpected encounters that also make cities the centers of innovation, as Jane Jacobs emphasized in her books. It isn’t just within one industry that these interactions are beneficial, but also when several industries interact within a city region, as, say, publishing, fashion, and (now only some) manufacturing have in the history of New York City.

So couldn’t each city region then replicate the entire set of industries? City regions aren’t large enough to provide everything they need; trade is critical to production. In the US, for instance, different city regions have specialized in different industries. I mentioned New York City, but Cincinnati was known for machine tools, Pittsburgh for steel, of course Detroit for cars, and we have had Silicon Valley, an outgrowth of the San Francisco economy, as the premiere example of a center of innovation. The idea is to have a large enough area to encompass all of the various niches of an economic ecosystem, and also one that is small enough to encourage a rich network of interactions.

In order for this weaving together of city regions to occur, the government has to create a transportation network. The Interstate Highway System served this purpose after World War II, as the railroad system did before (and as I have argued, probably will again sometime in the future). A communications network is also necessary, again generally either run by governments or supported by them.

Governments have historically also protected their territories economically and militarily in order to allow these regional production networks to grow to the point where they can compete globally. When governments bind together areas in this way, the regional economy becomes strong enough that global trade is actually increased. You need a world class production system before you can trade manufactured products, and governments have historically been a crucial builder of those regional economies — as I shall argue in my next post. On the other hand, when a country like the United States allows its manufacturing base to be exported, it will open up yawning trade deficits, and eventually, slide into poverty. The choice is ours.

There is no alternative: The government is part of the solution

The current conventional wisdom for many in the U.S. is that the less government is involved with the economy the better. But this is precisely the moment in history when more government is needed. Without government intervention, the recovery will continue to stagnate, the economy as a whole will remain off balance, and we won’t be able to meet the challenges facing the country.

I have been proposing a different way of looking at an economy than the traditional, neoclassic one. In my view, each industry fits into a wider system, as say trees or deer or bears fit into a wider forest ecosystem. In the same way, goods manufacturing, machinery industries, service industries, infrastructure, and the myriad other parts of a functioning society — including the health and education systems — have to work properly in order for the economy as a whole to function, with manufacturing functioning as the central sector. All industries are co-evolving, dynamically growing, concentrated within discrete geographical regions. And it is the responsibility of government to help orchestrate this interaction, or else it can turn into an ugly riot.

But at the root of the neoclassical world view is the idea that the economic system is self-regulating, that is, if the economy is pushed off course by “external” forces, then it will become stable by itself — without government interference. And yet we know that economies are constantly growing and changing — that is, they are not stable — and they are often under threat of recession and depression. That is why governments always have to be part of the solution. They are needed in order to support economic growth, maintain the right structure of the economy, and intervene when the economy goes bad.

FDR’s presidency is the perfect example of this. When he became president, Herbert Hoover had just spent several years trying to reverse the Great Depression with market-based solutions, but FDR championed a set of governmental policies that turned the country around. To deal with unemployment, FDR established the Works Progress Administration, or WPA, which was not only designed to employ one fully able member of each household in which no one could find work, but also to build up the country’s physical infrastructure. Building infrastructure is what governments do best. In fact, one could say that civilization started when the first governments constructed the irrigation and drainage systems that enabled agriculture to flourish. The United States, like every successful country, has a long and rich history of infrastructure building, without which the country would have very likely stayed poor. From canals like the Erie Canal before the Civil War, to the railroads after, from the dams that even conservative Republicans like Calvin Coolidge initiated, to the WPA that built libraries, schools, airports, roads, and other structures in virtually every town, to the Interstate Highway System championed by a Republican president, the United States has kept itself at the forefront of the global economy by making the building of transportation, energy, communications, water, education, and other systems the foundation of prosperity.

Partly as a result of his interventions into the economy, FDR was able to lead the nation into World War II by fundamentally transforming the economy to produce military equipment. At its height, one third of the country’s GDP was devoted to the war effort, with millions fighting overseas. That’s five trillion dollars in today’s economy. In other words, even assuming the continuation of a one trillion dollar military budget in the face of no wars of necessity, the economy has four trillion dollars left over to remake itself while providing for a comfortable standard of living for its inhabitants.

Instead of learning this lesson of history, however, our current political class seems determined to follow Herbert Hoover, not FDR. Meanwhile, the long-term domestic problems we face are worse than what FDR confronted. In the 1930s, the US was by far the leading manufacturing power and the top producer of oil; now the manufacturing sector is sinking fast, and not only do we import almost two-thirds of the crude oil we process, the global supply of oil is becoming harder to produce and is shrinking. In addition, we desperately need to eliminate the use of fossil fuels and transform agriculture and forest management in order to avoid the worst of global warming. The path forward is clear: we need an electric transportation system based on high-speed rail for long-distance travel, electric rail for freight, transit and small electric cars for intra-city movement, wind and solar power for electricity generation, recycling on a serious and massive scale, a densification of urban areas, and a more labor-intensive, localized, organic agricultural system. And these could provide the market for a revived manufacturing sector.

Only the government can build all of these systems in the time needed to both save the economy and save the environment. Incentives can go part of the way, but not fast or far enough. Taxing carbon or trading rights to carbon won’t solve global warming or decrease the use of oil as quickly as we need them to; lowering taxes or reducing the deficit won’t bring the manufacturing sector back. Government-as-builder does not mean government-as-warrior or government-as-Big-Brother. It is possible to have a strong government that is peaceful, democratic, and not beholden to our economic royalists, as we currently are. But maintaining democracy is never easy; the political system is no more a self-regulating system than is the economy. At least we can have a clear vision of where we are heading.

History doesn’t care if the political conversation of the United States won’t allow for talk about large-scale government intervention into the economy. The path to economic and ecological collapse is paved with “realistic” intentions. If the conservatives can be audacious enough to threaten policies that will further destroy the middle class and poor for the sake of the superwealthy, why can’t progressives draw on a rich American history, from before FDR and after, to rebuild a once mighty nation and help the rest of the planet move toward a sustainable future?