Total Factor Productivity (TFP) through innovation

[El_Machinae]

If I could ask a couple questions?

Apparently, if I am summarizing well, TFP is the transforming of inputs into outputs that are worth more than the inputs. We can get economic growth by using more inputs or by increasing the number of hours works, but that's not really 'wealth creating'. Whereas TFP seems to be actual wealth creation.

I found this graph.



Nondurable TFP seems to be the economic growth that's generated from innovations and ideas? When people come up with improvements to processes, or ways to improve products so that they're either cheaper or more valuable, is that nondurable TFP?

I think that graph means that the US has only been getting ~0.5% economic growth from 'ideas' for the last few decades. Or at least, that's nearly what nondurable TFP has been. Is there a reason why this stagnation is occurring? My interpretation is that, if we start to stabilise our natural inputs into the economy (population growth and resource extraction), we can only hope for 0.5% economic growth.
As an aside, so many reports on TFP use natural logs, when what I want is % of GDP!

Am I close? Am I using the right graph for what I think it's about?

 

[Mise]

I thought "durable goods" were things that lasted a long time, like TVs, cars, computers, etc, while "non-durable goods" were things that were consumed quickly or immediately, like TV programmes, petrol, internet access, etc. I don't know if it has a different meaning in the context of TFP though. The graph would suggest that process improvements to the manufacturing of TVs, cars etc are more significant than process improvements towards things like legal services and such.

 

[Cutlass]

Wealth is "created" either by gaining more inputs, or by having more outputs per level of input. I have made the argument in the past that, ultimately, all wealth comes from ideas. Because even with inputs, you need the idea that they are worth anything.

Ideas create new wealth. They allow the use of inputs to create ever greater values of outputs.

I'm not clear on what your graph is saying or what exactly it is that you are asking. Many "non-durable" services in the economy don't get a lot of productivity increase because labor doing the work by hand as it has always been done isn't easily replaced. And so the productivity is stagnant and reflected in your graph. Think of a machine that can change the sheets on a bed. That is an immensely complicated job for a machine, easy for a minimum wage maid.

 

[Integral]

I thought "durable goods" were things that lasted a long time, like TVs, cars, computers, etc, while "non-durable goods" were things that were consumed quickly or immediately, like TV programmes, petrol, internet access, etc. I don't know if it has a different meaning in the context of TFP though. The graph would suggest that process improvements to the manufacturing of TVs, cars etc are more significant than process improvements towards things like legal services and such.

Start with this, read it carefully and you're over halfway there. Your durables- nondurables distinction is correct and the graph is splitting TFP growth into those two sectors.

It's less about "stuff" vs "ideas" and more "stuff that lasts a while" vs "stuff that doesn't".

As an aside, so many reports on TFP use natural logs, when what I want is % of GDP!

TFP is typically calculated as a residual. With the specific functional forms we use to get that residual, the natural result is log(TFP).

You can also do fun things like decompose GDP growth into (1) labor component, (2) capital component and (3) TFP component. See Alwyn Young's "The Tyranny of Numbers" for a good look at measuring and applying TFP.

...how much detail do you want? I happen to be an expert on this subject.

 

[Terxpahseyton]

That's how I understand it:

First what Mise/Cutlass said regarding nondurable goods. That those goods still highly depend on human labor and by this productivity increases depend heavily on basic human abilities. As a consequence at some point only little room is left for further productivity increases.
Durable goods, all those gadgets you have and can use repeatedly, derive their main cost from production and production is very prone to efficiency increases by automatization, cheaper materials and production processes etc.

Or in other words: Durable goods rest heavily on flexible factors (factors which have to be created in the first place and hence can be changed at any time and in many directions), while nondurable goods rest heavily on fixed factors (human potential, which is more or less given by nature, not "created" and hence not so much within our control).

Another thing this graph shows is how - beginning in the 60s - durable goods lost importance in the overall economy (service economy). Which means that the economic sector with the most potential for wealth gains by improved efficiency (which can be viewed as the main cause of the economic flourishing in th 50s and 60s) has shrunk while the sector with little potential for wealth gains by bigger efficiency has grown.

 

[El_Machinae]

Ah, thank you. I was rethinking the thesis of the speaker who presented this term to me. His point was that there were limits to how much of a 'service' economy we could be, and that areas that require a large amount of human effort to create a product were going to continue to be a reason for economic stagnation.

I misunderstood his gist. I was hoping it would be the R&D contribution to economic growth. The metric I am interested in is "what do we measure to predict economic growth, if we assume a freezing of resource extraction?"

 

[warpus]

Am I close? Am I using the right graph for what I think it's about?

My opening question would be: Why is economic growth so important?

 

[Cutlass]

You also have to understand that many services are subject to increased productivity through innovation and capital equipment. Take, for example, transportation. Containerized freight made vast numbers of dock and warehouse workers obsolete. Computers made too many occupations to list more productive. The pager and cell phone did also. Bar code scanners. Developed nations, and many undeveloped as well, had great productivity improvements since the 90s just on the microcomputer and cell phone alone.

But you also have to understand the concept of substitutes. Substitutes aren't limited to plastics replacing metals. It also applies to capital replacing labor. And vice versa. When the price of capital is low versus the price of labor, firms will replace labor with capital, and productivity will go up. When this isn't good enough with current machines, there is great incentive to invent better machines. Innovation. For the innovation to be profitable, it has to make financial sense to buy equipment to replace labor.

But the opposite is also true. Where labor is cheap, it pays to stagnate or downgrade equipment and just use the cheap labor. This means that those workers will have stagnant or possibly even declining productivity. And depending on how large of a piece of the economy that is, it can effect the national stats.

Recall that all (developed and semideveloped) economies are a mix of the high and low productivity. The question becomes what is the mix between them?

So many occupations have low productivity because the process of automating them is too difficult, or because the labor is so cheap that there's no incentive to do so. But many others could have their productivity raised if there was just a bit more incentive.

This author argues that when production costs are high for some factors, firms offset that with innovation:

State Bad For Business? Not For Manufacturing
Surprise — Manufacturing Costs In Connecticut Among Lowest In Nation
July 17, 2011|By LEI CHEN, The Hartford Courant

Despite the conventional wisdom that the South is the best place in the United States for manufacturing, results from my recent study at the American Institute for Economic Research showed that eight of the top 10 states for cost efficient manufacturing are in the Northeast and the West.

Connecticut tops the list, tied for first place with Oregon and Iowa for being most cost-efficient. Other states in the top 10 cost-efficiency rankings are North Carolina, New York, Arizona, Massachusetts, Nevada, Colorado and Washington.

The study employs Data Envelopment Analysis, a completely data-driven method based on figures from the 2007 Economic Census. From an efficiency perspective, the analysis examined manufacturing costs that include production labor, non-production labor, capital, energy and materials.

States such as Connecticut — and its neighbors New York and Massachusetts — that are highly cost-efficient have manufacturing bases that replace high-cost resources needed for manufacturing with relatively low-cost ones. Connecticut firms have likely made sensible adjustments to the prevailing costs by economizing on more expensive resources, making fuller use of relatively cheaper alternatives and developing more efficient production methods.

.....

More here:
http://articles.courant.com/2011-07...-20110717_1_connecticut-tops-costs-production


More here: http://www.aier.org/component/conte...letin/2474-competitiveness-and-business-costs

 

[Integral]

My opening question would be: Why is economic growth so important?

Because economic growth gets you from here:


to here:


And frankly, I like the look of the second more than the first. In the first you have to worry about your poor people starving, in the second you have to worry about them being too fat.

 

[El_Machinae]

My opening question would be: Why is economic growth so important?

I have grand dreams, and I need a progressively wealthy society to realise them.