A few reasons strategy is hard

  • Strategy is dynamic. The best strategies emerge as complex decision environments evolve.
  • Strategy is rarely predictable. Because predictions are rooted in history, they work best in environments that aren’t changing, that aren’t going through a regime shift. 
  • Strategy is hard to practice. That is, it’s hard to backtest, especially if facing an important decision in a rapidly-changing decision space. 
  • The most interesting strategies are created, and aren’t known in advance. They aren’t announced. 
  • The most impactful strategies have long tails. And impacts that are felt far beyond their domains. 
  • Strategy is human. If people are purposive (they don’t have to be fully rational, even boundedly rational will do), then strategy is one of the most human activities. Even an AI will struggle because of the benefits of seeing “beyond the board”. 
  • In strategy, the most important ones emerge when the decision space is infinitely-dimensioned. People may struggle with that complexity, but the remarkable ones will find ways around it – not just to simplify it, but to transcend it. 
  • Strategy is granular. 

Data for dissertations August 28, 2017

35514 Quantifying the Size and Geographic Extent of CCTV’s Impact on Reducing Crime in Philadelphia, Pennsylvania, 2003-2013

35613 Expanded Assessment of the Consequences of Imprisonment for Employment in Maricopa County, Arizona from 2011-2012

36624 Los Angeles Metropolitan Area Surveys [LAMAS] 5, 1972

36645 Afrobarometer Round 6: The Quality of Democracy and Governance in Burundi, 2014

36679 Afrobarometer Round 6: The Quality of Democracy and Governance in Cape Verde, 2014

New data from ICPSR

36760 Annual Survey of Jails, 2015 http://doi.org/10.3886/ICPSR36760.v1

36800 Research on Pathways to Desistance [Maricopa County, AZ and Philadelphia County, PA]: Subject Measures – Scales, 2000-2010 http://doi.org/10.3886/ICPSR36800.v1

36825 Child Care and Development Fund Administrative Data, Federal Fiscal Year 2014 http://doi.org/10.3886/ICPSR36825.v1

36842 Health Reform Monitoring Survey, Third Quarter 2016 http://doi.org/10.3886/ICPSR36842.v1

Policy Design and Practice

I’ve joined the Scientific Board of Policy Design and Practice, a new journal from Taylor & Francis that will begin publication in 2018. The journal is sponsored by the Asian Development Bank and the Lee Kuan Yew School of Public Policy of the National University of Singapore. Michael Howlett and M Ramesh will co-edit the journal. I’ll post more information along the way about the journal.

Messiness is generative

I’m going to write for a bit on innovation and design and research and a few other topics. It fits with some work I’m doing on innovation in complex learning organizations, but I won’t focus on the specifics much at all in these entries. These are just sketches of some ideas I’m enjoying reading about.

For instance, I’ve been reading about ethnographic studies of graphic designers. One item that struck a note is the idea that they work in teams, and the social nature of the design practice is a big part of what makes for success. The groups brief each other in open studios where people can hear and see conversations, often carried out in open conflict. The detritus of design is left scattered about the room – drawing boards, sketches, all kinds of papers. Nothing is hidden – things carved in foam are jumbled together on desks along with forgotten prototypes.

The point is that messiness is generative. Seeing a forgotten prototype for an early stage in a project may remind a designer of an important tradeoff made at that point; maybe it even helps get around a newly-discovered juncture.

What if research was carried out in the same environment? If we left our solo offices to sit in bullpens surrounded by other researchers and grad students? If we published (if only electronically, as a working paper on a server like the arXiv) every failed analysis? Every conjecture or hypothesis we considered worth extended consideration?

Certainly it would be messy. It would open us up to enhanced criticism. But it would also be generative.