Archive for October, 2018

On the recent Pew poll contrasting differences as to the “very big” problems we face today

October 20, 2018

An online news item appearing on 15 October 2018 proclaims that “Americans don’t just disagree on the issues. They disagree on what the issues are.” The article, by Dylan Scott on the Vox website, reports on a poll conducted by the Pew Research Center, involving registered voters in the U.S., between 24 September and 7 October. Polarizing disagreement is a recurring theme in the world, and keeping the tension up sells ads, so it is not surprising to see the emphasis in both the article and in the Pew report on major differences in people’s perceptions of what counts as a “very big” problem in the U.S. today. But a closer look at the data gives hope for finding ways to communicate across barriers that may look more significant than they actually are.

There’s no mention in the article of the sampling error, uncertainty, or confidence level, but the Pew site indicates that, overall, sampling error is 1.5%. But the Vox article mentions only the total sample size and fails to say that the registered voter portion of the respondents is smaller by a couple of thousand. Further, the sampling error jumps up to 2.6% for respondents indicating support for a Republican candidate, and to 2.3% for respondents supporting a Democrat. Again, the differences being played up are quite large, so there’s little risk of making too much out of a small difference. It’s good to know just how much of a difference makes a difference, though.

That said, neither Pew nor the Vox story mentions the very strong agreement between the different groups supporting opposing party candidates when the focus is on the relative magnitudes of agreement on aligned issues. Survey research typically focuses, of course, on percentages of responses to individual questions. Only measurement geeks like me wonder whether questions addressing a common theme could be related in a way that might convey more information. My curiosity was piqued, even though it is impossible to properly evaluate a model of this kind from the mere summary percentages. I knew if I found any correspondences they might just be accidents or coincidences, but I wanted to see what would happen.

So I typed up the text of the 18 issues concerning the seriousness of the problems being confronted in the US today, along with the percentages of registered voters saying each is a “very big” problem today. I put it all into SPSS and made a few technical checks to see if any major problems of interpretation would emerge from the nonlinear and ordinal percentages. The plots and correlations I did indicated that the same general results could be inferred from both the Pew percentages and their logit transformations.

While I was looking at a scatter plot of the Republican vs Democrat agreement percentages I noticed something interesting. I had been wondering if perhaps the striking differences in the groups’ willingness to say problems were serious might be a matter of relative emphases. Might the Republican supporters be less willing to find anything a big problem, but to nonetheless rank the issues in the same order as the Democrat supporters? This is, after all, exactly the kind of pattern commonly found in data from various surveys, assessments, and tests. No matter whether a respondent scores low overall, or scores high, the relative order of things stays the same.

Now, this is true in the kind of data I work with because considerable care is invested in composing questions that are intended to hang together like that. The idea is to deliberately vary the agreeability or difficulty of the questions so they all tap the same basic construct and demonstrably measure the same thing. When these kind of data are obtained, different questions measuring the same thing can be asked of different people without compromising the unit of measurement. That is, each different examinee or respondent can answer a unique set of questions and still have a measure comparable with anyone else’s. Like I said, this does not just happen by itself, but has to come about through a careful process of design and calibration. But the basic principles are well-established as being of longstanding and proven value across wide areas of research and practice.

So I was wondering if there might be one or more subsets of questions in the Pew data that would define the same problem magnitude dimension for supporters of both Republican and Democratic candidates. And as soon as I looked at the scatterplot of the percentages from the two groups, I saw that yes, indeed, there appeared to be four groups of issues that lined up along shared slopes. A color-coded version of that plot is in Figure 1.

The one statistical inference problem that emerged in examining these ordinal data concerns the yellow dot that is lowest and furthest to the left. At 8% agreement from the Republican supporters it was pulled away from the linear relation further than the other correspondences. When transformed into a log-odds unit, that single problematic difference lines up well with the other yellow dots further to the right.

The identity line in the figure shows where exact agreement between the two groups would be. That line marks out the connection between the same percentages of respondents agreeing an issue is a “very big” problem. We can see that the three green dots very nearly fall on that identity line. Just below them is a row of blue dots almost parallel with the identity line. Then there’s a third row of yellow dots further down, indicating more absolute disagreement between the two groups on these issues, but also showing a quite strong agreement as to their relative magnitudes within that group. Finally, there is another, red, line of dots in the lower right corner of the figure that marks out a more extreme range of absolute disagreement, but is also quite parallel to the identity line.

Fisher2018PewFig1

Figure 1 Initial plot of Republican vs Democrat Percentages agreement as to “Very Big” problems

Figures 2-5 below illustrate each of these groups of issues separately, giving further information on the problems and showing the regression lines and correlations for each contrast. The same colors have been retained to aid in seeing which groups of issues in Figure 1 are being shown.

The four areas of problems seem to me to correspond to issues of perceived major threats (Figure 2), accountability and access issues (Figure 3), equal opportunity issues (Figure 4), and systemic problems (Figure 5). Each of these content areas could be explored conceptually and qualitatively to assess whether some initial sense of a measured construct can be formed. If the by-person individual response data could be analyzed for fit to a proper measurement model, a much better job of determining the presence of invariant structure could be done.

But even without undertaking that work, these results already suggest a basis for productive conversations between the supposedly polarized groups. To start from the low-hanging fruit, the three problems the two groups agree on to within a couple of sampling errors (Figure 2) present topics of common agreement. Both Democrats and Republicans identify violent crime, the federal budget deficit, and drug addiction as matters of equally shared concern. The point is not that these are the highest rated problems for either group, but, rather, that they agree within the limits of statistical precision as to the extent that these are “very big” problems. It may be that setting shared priorities for addressing these problems could ground new relationships in that experience of having accomplished something productive together.

This new approach to building social capital might then proceed by taking up progressively more difficult areas of disagreement as to what “very big” problems are. Even though Republicans rate each area as less likely to be a “very big” problem, within each of the four groups of issues, they agree with Democrats as to their relative magnitudes. News like this might not sell a lot of ads, but it does offer hope for finding new ways of approaching relationships and crossing divides.

Fisher2018PewFig2

Figure 2.Republican vs Democrat areas of agreement as to “Very Big” problems

Fisher2018PewFig3

Figure 3 Republican vs Democrat areas of some disagreement as to “Very Big” problems

Fisher2018PewFig4

Figure 4 Republican vs Democrat areas of marked disagreement as to “Very Big” problems

Fisher2018PewFig5

Figure 5 Republican vs Democrat areas of fundamental disagreement as to “Very Big” problems

Creative Commons License
LivingCapitalMetrics Blog by William P. Fisher, Jr., Ph.D. is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Based on a work at livingcapitalmetrics.wordpress.com.
Permissions beyond the scope of this license may be available at http://www.livingcapitalmetrics.com.

Why economic growth can and inevitably will be green

October 1, 2018

So, approaching matters once again from yet another point of view, we have Jason Hickel explaining a couple of weeks ago “Why Growth Can’t Be Green.” This article provides yet another example of how the problem is the problem. That is, the way we define problems sets up particular kinds of solutions in advance, and sometimes, as Einstein famously pointed out, problems cannot be solved from within the same conceptual framework that gave rise to them. I’ve expanded on this theme in a number of previous posts, for instance, here.

Hickel takes up the apparent impossibility of aligning economic growth with environmental values. He speaks directly to what he calls the rebound effect, the way that “improvements in resource efficiency drive down prices and cause demand to rise—thus canceling out some of the gains.” But that rebound can happen only as long as the economy remains defined and limited by the alignment of manufactured capital and finance, ignoring the largely unexamined and unconsidered possibility that human, social, and natural capital could be measured well enough to be also aligned with finance.

Hence, as I say, the problem is the problem. Broadening one’s conceptualization of the problem opens up new opportunities that otherwise never come into view.

The Hickel article’s entire focus is then on top-down policy impositions like taxes or a Genuine Progress Index. These presume human, social, and natural capital can only ever exist in dead formations that have to be micromanaged and concretely manipulated, and that efficient markets bringing them to life are inherently and literally unthinkable. (See a short article here for an explanation of the difference between dead and living capital. There’s a lot more where that came from, as is apparent in the previous posts here in this blog.)

The situation could be vastly different than what Hickel imagines. If we could own, buy, and sell products in efficient markets we could reward the production of human, social, and environmental value. In that scenario, when improvements in environmental resource efficiency are obtained, demand for that new environmental value will rise and its price will go down, not the resource’s price.

We ought to be creative enough to figure out how to configure markets so that prices for environmental resources (oil, farmland, metals, etc.) can stay constant or fall without increasing demand for them, as could happen if that demand is counterbalanced and absorbed by rising human, social, and environmental quality capital values.

The question is how to absorb the rebound effect in other forms of capital that grow in demand while holding demand for the natural resource base in check. The vital conceptual distinction is between socialistic centralized planning and control of actual physical entities (people, communities, the environment, and manufactured items), on the one hand, and capitalistic decentralized distributed network effects on abstract transferable representations, on the other. Everyone defaults to the socialist scenario without ever considering there might be a whole other arena in which fruitful possibilities might be imagined.

What if, for instance, we could harness the profit motive to promote growth in genuine human, social, and environmental value? What if we were able to achieve qualitatively meaningful increases in authentic wealth that were economically contingent on reduced natural resource consumption? What if the financial and substantive value profits that could be had meant that resource consumption could be reduced by the same kinds of factors as have been realized in the context of Moore’s Law? What if a human economics of genuine value could actually result in humanity being able to adjust the global thermostat up or down in small increments by efficiently rewarding just the right combinations of policies and practices at the right times and places in the right volumes?

The only way that could ever happen is if people are motivated to do the right thing for the earth and for humanity because it is the right thing for them and their families. They have to be able to own their personal shares of their personal stocks of human, social, and natural capital. They have to be able to profit from investments in their own and others’ shares. They will not act on behalf of the earth and humanity only because it is the right thing to do. There has to be evidence and explanations of how everyone is fairly held accountable to the same standards, and has the same opportunities for profit and loss as anyone else. Then, and only then, it seems, will human, social, and environmental value become communicable in a viral contagion of good will.

Socialism has been conclusively proven unworkable, for people, communities, and the environment, as well as financially. But a human, social, and natural capitalism has hardly even been articulated, much less tried out. How do we make human, social, and natural capital fungible? How might the economy transcend its traditional boundaries and expand itself beyond the existing alignment of manufactured capital and finance?

It’s an incredibly complex proposal, but also seems like such a simple thing. The manufactured capital economy uses the common language of good measurement to improve quality, to simplify management communications, and to lower transaction costs in efficient markets. So what should we do if we want to correct the imbalanced negative impacts on people, communities, and the environment created by the misplaced emphasis on aligning only manufactured capital and financial capital?

As has been repeatedly proposed for years in this blog, maybe we should use the manufactured capital markets as a model and use good measurement to improve the quality of human, social, and environmental capital, to simplify communications and management, to lower transaction costs, and to align the genuine human, social, and environmental value created with financial value in efficient markets.

Of course, grasping that as viable, feasible, and desirable requires understanding that substantively meaningful precision measurement is something quite different from what usually passes for quantification. And that is an entirely different story, though one taken up repeatedly in previous entries in this blog, of course….

 

Creative Commons License
LivingCapitalMetrics Blog by William P. Fisher, Jr., Ph.D. is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Based on a work at livingcapitalmetrics.wordpress.com.
Permissions beyond the scope of this license may be available at http://www.livingcapitalmetrics.com.

 

Common currencies for the exchange of human, social, and environmental value

October 1, 2018

I was just now reading an article (see link below) that underscores my conviction that the secure ledger platforms are the real message, and that cryptocurrencies are intuitions of the need for fungible expressions of scientifically and meaningfully measured human, social, and environmental value.

We absolutely need efficient markets to be able to buy, invest in, profit from, and scale up UN SDG sustainability products like carbon sequestration, reduced violence, and improved literacy rates. Linking real value with common product definitions and financial value will be complex, multilevel, and multifaceted, but it can be done. It will be expensive to create efficient markets for economically self-sustaining sustainability impacts, but doing so will pay returns many times greater than what’s invested. And neither doing nothing nor continuing as we are can stand as viable options.

When the right combination is hit, watch out! This event will reveal the Internet’s true purpose. Some will say it fulfills humanity’s destiny as stewards of the earth.

The economic transformation that follows will make everything that’s happened to date pale in comparison. It will become impossible to generate financial profits while destroying human, social, or environmental value. The alignment of financial and genuine wealth will even out monetary flows in such a way as to make a guaranteed minimum income nothing more than an obvious and inarguable consequence of economic reality. Critical engagement will lead sometimes to systemic improvements, sometimes to clear refutation of the critic, sometimes to a recognized need for more information, and more often to a new tolerance and respect for differences of opinion as our capacities to learn from one another grow.

All that to say I remain unconvinced by stories like this one that seem blind to the poetry, beauty, meaning, and power of taking living language as a model for adaptive, intelligent institutions capable of improving the human condition:

She prosecuted Bitcoin crimes. Now she’s a major cryptocurrency investor. – FORTUNE

For more on these matters, see previous posts here, my web site, my publications, etc.


LivingCapitalMetrics Blog by William P. Fisher, Jr., Ph.D. is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Based on a work at livingcapitalmetrics.wordpress.com.
Permissions beyond the scope of this license may be available at http://www.livingcapitalmetrics.com.