… and a star to steer her by?New ICT Governance and the Resilience

PrinciplesPierre de Vries

Sr. Adjunct Fellow, Silicon Flatirons Center

Presentation at Silicon Flatirons 2010 Annual Conference“The Digital Broadband Migration: Examining the Internet's Ecosystem”

1 February 2010

“Sea-Fever”

• John Masefield (1878-1967), English Poet Laureate, 1930-1967

I must down to the seas again, to the lonely sea and the sky,And all I ask is a tall ship and a star to steer her by,And the wheel's kick and the wind's song and the white sail's

shaking,And a grey mist on the sea's face, and a grey dawn breaking.

• Reminded as thinking of governance, “kubernan”• Tempting metaphor, but wrong– No single ship of state, no single course to agreed

destination– Will propose alternative approach

Yellow slides not shown in live presentation

New Models of Governance

Introduction

• Constant flux of new ideas for regulating ICT, internet/web

• Jonathan Sallet’s hypothesis: new models of governance are emerging– Led to Silicon Flatirons “

New Models of Governance” meta-program– Analyzed bottoms-up policy discussions in the

program, looking for top-down patterns

Sample Changes in Governance

Topic Context Object Agent Method

Cybersecurity

New threat

No counter devised

Cross-border nature

“cyberspace”

Create Nat’l Office for C’space; Cybersecurity Directorate in NSC

Increased Federal role compared to off-line

New regulations for industrial control systems

Mandate strong authentication

Network neutrality

Internet overtakes telecom network

“broadband”

“neutrality”

“network management”

Self-regulatory orgs

Shift competition responsibility to FTC

Crowdsourcing

Four Freedoms

Draft legislation

Use adjudicative powers, common law reasoning

Software patents

¼ of all issue; many patents per product; more intangible; rapid change

“software patent”

Courts vs. legislature

End the PTO monopolyDisallow software patents

«Plus ça change, plus c'est la même chose»

(?)

19th century journalist Jean-Baptiste Alphonse Karr

But:

• Nothing obvious emerged! Detail changes, no new models

• And yet… something’s going on – are we understanding it correctly? Either:– nothing’s changed– something’s changed

• Existing principles work; we’re done, move along now• New principles needed/emerging, but can’t recognize

• Guess: new underlying philosophy emerging, not overarching model

• Most obvious changes: context of governance

Changes

• Modularity• Convergence• Decentralization

• Third sector• Tempo• Scale

Cyclical

Step change

Context changes: Cycles

• Modularity– Public interfaces, standards, competition– In industry structure as well as technology

• Convergence– Old distinctions blur: silos inapplicable

• Decentralization– From centralized/hierarchical to disintermediated dumb network,

smarts at the edge• BUT

– Technology isn’t destiny: ~ technology does not lead inescapably to a ~ industry structure• Proprietary integration will hide modularity• New categories will come - “human rage to classify” – layers, new industries• Rise of online intermediaries: Google, Facebook, ISPs {cf. Paul Ohm}

• Cycles, not inevitable, one-way, monotonic development– new configuration

Context changes: Step changes

• Third sector– Rise of NGOs, non-profits and civil society: “self-governing private non-profit

organizations, pursuing public purposes outside the formal apparatus of the state” (Salamon)

– Telecom-Internet: standards from ITU to IETF/W3C/IEEE• Tempo

– William Scheuerman’s social acceleration of time• tech innovation, patterns of social change (family, workplace), everyday life via new

means of high-speed communication and transportation– Institutions not built to cope with pace of life – rise of agencies, power shift to

executive• Scale

– Huge number & diversity of apps/devices/services per person– Data aggregation/mining

• Characteristics– accelerated by internet/ICT– even if it doesn’t keep growing, unlikely to shrink: qualitative change

Complex Adaptive Systems

Cycles & phase changes

Incomplete knowledge

Cross-linked hierarchy

Novelty & surprise

Complex Adaptive Systems (CAS)

• CAS: a collection of interacting, adaptive agents– e.g. human body, ecosystems, economy

• Cyclical & step changes– similar to ecosystem cycles and state transitions: growth, maturity, collapse,

reorganization

• Incomplete knowledge– Deep uncertainty about how system works, what state it’s in, what the problem is, what

counts as a solution

• Hierarchy and cross-linking– Layers model, with linkages: security detection in network transport and in applications– Concurrent changes at different scales: video services, player plug-ins, transports

• Novelty and surprise– Rise of P2P traffic, Open Source/Linux– Unintended consequences: TA96 supposed to increase competition, but reduced it– Robust-yet-fragile behavior: e.g. subtle inconsistencies in protocol implementation or

router configuration (Pakistan YouTube)

So What?

• Internet ≠ an Ecosystem– but both have same underlying dynamics: CAS– cf. Whale ≠ Elephant: both large mammals

• Response– Take lessons from Complex Adaptive Systems theory

• Long literature: systems 50-60 yrs, complex systems 20-30 yrs, managed adaptive systems 10-20 yrs

– Focus on managed ecosystems, not autonomous closed systems (not just “nature red in tooth & claw”)

– Capture best practices in the “Resilience Principles”

The Resilience Principles

Flexibility

Delegation

Big Picture

Diversity

Why Resilience?

• Policy Imperatives: want innovation and stability BUT– Innovation is disruptive– Striving for immutability sets up the conditions for a

catastrophic collapse, e.g. fire suppression, protecting fading industries

• Resilience: “maintaining structure and function in spite of experiencing disturbances”

• Top-level rule of thumb for dealing with complexity and contradiction

• Not efficiency: cf. choosing a solution– Robust/resilient: performs reasonably well, compared to the

alternatives, over a wide range of plausible scenarios– Optimal/efficient: performs best in the most plausible scenario– Cf. Cheney “One Percent doctrine”: treat a 1% chance as a

certainty

Flexibility

• Long-term prediction is impossible; knowledge is inadequate; system adapts faster than controls can change; different parts in different stages

• “Neutral, open-ended policies. Determine ends, not means. Describe and justify the outcomes sought, not the methods to be used to achieve them”– Use principles rather than rules, e.g. solve ex post rather than guess ex

ante• Mechanisms/Examples

– Flexible-use radio licenses– Common law reasoning (Sallet, Weiser) in Network Neutrality (NN)

• Find the facts; ask if they are the same as or different from previous facts while isolating the difference between the facts that matter from those that do not; recognize the larger principle that arises from case-by-case decisions, and then, finally, ask whether the larger principle, as used in the past, still makes sense given societal changes

• Comcast NN case – styled as adjudication, but didn’t use ALJs for fact-finding– Experiment

Delegation

• Regulator only has limited control– close direct management often harmful, e.g. flood control,

government-protected rates for intl call termination; • “Harness discretion of local experts. Most problems should

be solved by the market and society, not by government. Government's role is to provide proper incentives and guidance, and to intervene to solve critical shortcomings.”

• Network neutrality example: self-regulatory orgs– Silicon Flatirons network management (Aug 2008): develop

norms and best practices, review net mgmt techniques, provide advisory opinions, enforce standards

– Verizon/Google TAGs (Jan 2010): develop best practices, act as a forum for dispute resolution, issue advisory opinions, and coordinate with standards bodies

Big Picture

• Emergent properties: overall behavior can’t be predicted from sub-systems; cannot optimize piecewise; narrow focus reduces robustness – e.g. protecting local manufacturing, 1950’s template for TDD accessibility

• “Take a broad view of the problem and solution space. Prefer generic to sector-, technology-, or industry-specific legislation.”– Particularly useful when objects of governance are changing (e.g.

periods of convergence)– Moving from stable/compartmentalized industry structure requires

new tools to account for feedback, non-linearities• Example: Simulation/modeling

– Agent-based modeling, genetic algorithms, systems dynamics– Way to grasp big picture, experiment with solutions– NN example: J Bauer and K DeMaagd (2008): genetic programming

techniques to model the co-evolution of platform operators, content providers, and consumers subject to specific policy rules governing the interactions

Diversity

• Diversity increases resilience– Biodiversity; part of value of competition

• Needs to be maintained in socio-economic system: anti-trust– Reduction in diversity amounts to an efficiency/resilience trade-off;

the resulting system is more efficient (standards, stability), but less resistant to shocks

• “Allow and support multiple solutions to policy problems. Encourage competition and market entry.”

• Examples– SMP analysis of NN: “European” approach

• Precedent in radio auctions: rules to preclude concetration– Recruiting citizenry to policy making process

• New agent in the governance mix – Crowdsourcing: grassroots organizing SaveTheInternet.com, FCC’s soliciting input OpenInternet.gov

• But difficulties: theater; interpretation of input; capture (astroturf -> cyberturf)

Conclusions

• Underlying all the point changes we’re seeing is a shift to complex systems thinking– in methods– driven by changed characteristics

• Ecosystem management provides a framework for defining and implementing new models: the Resilience Principles– Doesn’t mean previous approaches were wrong –

explains why what some were right, and guides choices for new ones to come

– Even if ICT complexity isn’t without precedent, we now have tools we did not have before

“Well, what do you know about that! These forty years now, I've been speaking in prose without knowing it! How grateful am I to you for teaching me that!”

Monsieur Jourdain in Moliere's The Bourgeois Gentleman (1670)