Brewed Fresh Daily

6 Responses to “Interesting guidance on entrepreneurial activity”

1. Justin Balck Says:
   October 30th, 2008 at 4:11 pm

   The most discouraging statistic was the drop in NEO VC investing from 30 in 1997 to 37th in 2006. And this period even coincided with two booms (tech, mortgage) and one bust (tech).

   Who is to be held accountable for this decade of failure? Let’s turn the tax and foundation money spigot off until Dr. Shane’s prescribed self analysis is implemented.

2. John Ettorre Says:
   October 30th, 2008 at 7:52 pm

   I think poor Dr. Shane needs to get his face out of the books and his piles of statistics for just a moment and do some old fashioned reporting, meaning walking and talking and especially listening to real businesspeople. I think he’d find a very different picture than he’s been painting for the last year or so. His conclusions simply don’t ring true with anyone who’s spent a lot of time watching this region and/or writing about what’s really going on in the business community. His curious focus on the glass half empty to the exclusion of the other half is, I’m afraid, more a reflection on the state of academic research than it is on the underlying dynamics upon which he purports to describe.

3. J Murray Says:
   October 31st, 2008 at 1:16 pm

   John, no, I think your conclusion is incorrect. Scott focuses in his paper on what can be proved from the evidence, and clearly considers “the popular wisdom” on various subjects. The distinction between what can be scientifically proven and what is believed or commonly accepted is a critical one in policy debates and, I’m afraid, one that is often scoffed at, overlooked, or disregarded by people who want to leap into action based on beliefs that either are not yet supported by evidence, or will never be supported by evidence.

   As I have written, it is a common blind spot of us all to believe the evidence of our eyes, but not to understand that the evidence that is presented to each of us individually may be unrepresentative of the entire set of evidence that is available for scientific inquiry. Let’s explore a specific here: Scott’s conclusion that attracting immigrants has a weak correlation with economic growth.

   First of all, Scott is comparing the variable “immigration” against other variables and ranking them by effect to determine those with the higher correlation with desired outcomes–the biggest bang for the buck. In this analyis–to simplify–each variable would be an X in a two-dimensional statistical analysis with result Y, which would be economic growth or one of its proxies job creation, wealth creation, etc. So picture an XY graph with an alternating series of X variables, each of which causes some shift in the desired outcome Y. What Scott is saying is that there are other X variables that produce a greater positive change in the desired outcome than attracting immigrants.

   In a world of limited resources, which is the world in which the Fund for Our Economic Future operates, choices have to be made about where to allocate resources. If the decision is to allocate resources to those with the biggest bang for the buck, attracting immigrants would not be the place to put those limited resources.

   This is not to say that attracting immigrants is a bad thing, or that it would not be a good thing to attract immigrants. Only that the methodology of the analysis Scott conducted would lead to the conclusion that the FFEF with its limited resources should not engage in that activity until it has first engaged in all other activities with a demonstrably greater impact on desired outcomes.

   Now, one might ask, why did Scott’s analysis determine that immigration has a weak correlation with entrepreneur-driven economic growth? That is an interesting question to explore. To understand the answer, you would have to understand the data sets that were available to Scott on which his analysis were baseds.

   I don’t have those, but I do have one hypothesis: some 30 years ago, U.S. immigration policy was deliberately changed to encourage family unity, and this resulted in a pool of immigrants that was composed only party of entrepreneurs, and of many other non-entrepreneurial people. Under this policy change, when one member of a family immigrated to the U.S., that person’s relatives were subsequently favored. Let us suppose that the person who came first was entrepreneurial, but the ones who followed were not. The entrepreneur was followed by his elderly parents, wastrel siblings, and extended family members who wanted to ride the entrepreneur’s gravy train. If this hypothesis were supported by the evidence, it would explain why Scott’s analysis showed what it showed: the data set would include some entrepreneurs, and many family members who were not entrepreneurs who watered down the effect of immigrant entrepreneurs on economic growth.

   I strongly believe that immigrants drive entrepreneurship, and have lots of personal, anecdotal evidence to that effect in my life and investments. That doesn’t, however, mean that the entire pool of immigrants is entrepreneurial; I only meet the ones who are. This is, in a personal way, the distinction I made at the beginning of the post between what one person sees and believes, and what the entirety of the evidence supports.

   My conclusion from this study, regarding immigration, is that the question I posed should be studied and if the hypothesis is validated, immigration policy should be changed to favor entrepreneurs over family members. That public policy initiative would be, of course, beyond the scope of the FFEF to undertake alone.

4. Ed Morrison Says:
   October 31st, 2008 at 4:05 pm

   It may be that we are reaching the limits of regression analysis in understanding the complex system dynamics of entrepreneurship in a regional economy.

   Shane relies on this tool to reach his policy conclusion, yet regression analysis is quite limited as a policy instrument. It’s a little bit like looking through a straw. With it, we can only see a limited range.

   So, for example, regression analysis assumes that there is a laundry list of factors associated with the given phenomenon we are trying to understand. The method assumes that causality is largely linear, that causality runs one way, and that each factor operates independently. Regression also assumes that the influence of each factor is stable over time.

   Regressions do not provide much insight into the underlying system dynamics. If we are interested in dramatically improving the performance of a complex system — – finding the 20% that generates 80% — then we need to build more sophisticated models of regional relationships.

   Here again, regression analysis is of limited value, since correlation is not adequate to help us understand how a complex system works. Systems thinking assumes that factors are interdependent, and that the character of these interdependencies can vary over time.

   These assumptions lead to greater complexity and sophistication. Exploring these connections opens the door to more practical insights about the leverage points that can improve dramatically the performance of a complex system.

5. J Murray Says:
   October 31st, 2008 at 4:46 pm

   Ed, I think that’s a good point. The challenge we have in considering complex multivariate systems is the absence of, or complexity of the available tools. A second problem is engaging people of low scientific literacy in understanding what scientific analysis can and cannot do. A third challenge is beliefs that defy attempts at educating people about the results and limitations of scientific inquiry.

   I’m a big believer in thinking about these things as systems, and that’s because I have spent a lot of time studying evolutionary biology and the many corollary inquiries it leads to. You can see that type of thinking in the NorTech Early Stage Capital Task Force reports, which recommend increasing the regional supply of seed capital, venture capital, and late stage capital to support companies at different stages–when a narrower view might have been just to focus on the apparently (at that time) inadequate supply of preseed capital.

6. Ed Morrison Says:
   October 31st, 2008 at 7:46 pm

   Jonathan:

   Economic thinking is moving toward complex adaptive systems and more sophisticated tools for understanding how to improve the productivity of these complex systems. While I have no doubt that Shane has done a good job in his work, I question the approach.

   Biological models, which focus on open networks, adaptation and signaling pathways, among other features, are beginning to replace outmoded linear assumptions and belief systems.

   Complex adaptive systems, from ecosystems to large corporations, share a number of common characteristics, including the ability to adapt to changing environments. Economics is moving gradually to adopt these new models. (See The Origin of Wealth.)

   The Santa Fe Institute has been pioneering the development of new thinking. The essential factor to innovation is finding the proper mix between order and complexity, between infrastructure and information.

   This move toward complex adaptive systems has significant implications for public policy at all levels. Effective policy must also balance structure and the inherent instability of adaptation.

   One of the most successful examples is the Small Business Innovation Research Grants initiative. At Purdue we are adapting this model within the framework of Open Source Economic Development. We have managed a $15 million investment following these principles. We operate with a staff of two full time people.

   We are finding that small investments, carefully placed based on an underlying analysis of a system, can trigger much larger investment commitments.

   This “link and leverage” strategy provides much more immediate and precise feedback on value creation than a regression analysis, which relies on high level, relatively abstract data; linear assumptions on causality; and static assumptions about “independence” and “dependence”.