Core Principles and Applications of Upstream Thinking

This book synthesizes the core principles of “upstream thinking,” a framework for preventing problems rather than reacting to them. The central thesis is that society is disproportionately focused on downstream responses—addressing crises, emergencies, and failures after they occur. An upstream approach, conversely, involves proactively identifying and dismantling the systems that cause these problems in the first place. This shift is impeded by three primary barriers: Problem Blindness, the failure to see a problem or the belief that it is inevitable; Lack of Ownership, a mindset where those capable of fixing a problem believe it is not their responsibility; and Tunneling, a state of scarcity (of time, money, or bandwidth) that forces short-term, reactive thinking and precludes long-term planning. Successful upstream interventions require leaders to unite diverse teams, identify high-leverage points within complex systems, establish early warning signals, and secure funding for outcomes that are often invisible—the absence of problems. The analysis reveals that effective upstream work is not about finding a single “magic pill” solution but about creating data-rich “scoreboards” that enable continuous learning and systems-level change.

——————————————————————————–

1. The Upstream Philosophy: Prevention Over Reaction

The core concept of upstream thinking is captured in a public health parable: two friends rescuing an endless stream of drowning children from a river, until one goes upstream “to tackle the guy who’s throwing all these kids in the water.” This metaphor distinguishes between downstream actions, which react to problems, and upstream efforts, which aim to prevent them.

Defining Upstream vs. Downstream Action

• Downstream Action: Reactive, tangible, and focused on restoration. Examples include a call center representative resolving a customer complaint, a doctor performing bypass surgery, or a police officer making an arrest after a crime. These actions are often demanded by circumstance.
• Upstream Action: Proactive, preventative, and focused on systems change. It involves “systems thinking” to systematically reduce the harm caused by problems. Examples include redesigning a website so customers don’t need to call for help, promoting policies that support healthy lifestyles to prevent heart disease, or creating community opportunities that deter crime. These efforts are chosen, not demanded.

The further one moves upstream, the more complex, ambiguous, and slower the solutions become, but the potential for massive and long-lasting good increases significantly. An intervention can exist at many points along a spectrum; for example, swim lessons are further upstream than life preservers in preventing drowning.

The Case of Expedia: A Model for Upstream Intervention

The travel website Expedia provides a clear illustration of a successful upstream intervention.

• The Downstream Problem: In 2012, 58 out of every 100 Expedia customers placed a support call after booking. The top reason, accounting for 20 million calls annually at a cost of roughly $100 million, was to request a copy of their itinerary.
• The Downstream Mindset: The call center was managed for efficiency—minimizing call time—rather than questioning why the calls were necessary.
• The Upstream Shift: A “war room” was created with a mandate to “Save customers from needing to call us.” They analyzed the root causes of the calls.
• Upstream Solutions: For the itinerary issue, they implemented simple fixes: adding an automated voice-response option, changing email protocols to avoid spam filters, and creating an online self-service tool.
• The Result: The 20 million itinerary-related calls were virtually eliminated. The overall percentage of customers needing to call for support dropped from 58% to approximately 15%. This success was achieved by integrating the work of different teams (product, tech, support) to solve a problem that no single group “owned.”

The Asymmetry of Attention: Why Society Favors Reaction

Despite the clear benefits of prevention, societal efforts are overwhelmingly skewed toward reaction.

• Tangibility and Measurement: Downstream work is more tangible and easier to measure. A police officer who writes a stack of tickets has a visible output, while an officer whose presence on a dangerous corner prevents accidents has invisible victims and victories written only in declining data.
• Funding and Resources: We spend billions to recover from disasters like hurricanes and earthquakes, while disaster preparedness is “perpetually starved for resources.” The U.S. healthcare system, a $3.5 trillion industry, is designed almost exclusively for reaction, functioning like a giant “Undo button” for ailments rather than a system for creating health.
• Heroism: Society celebrates the rescue, the recovery, and the response. Upstream work creates a quieter breed of hero, one “actively fighting for a world in which rescues are no longer required.”

Case Study: Healthcare Spending in the U.S. vs. Norway

The contrast between U.S. and Norwegian healthcare spending illustrates the consequences of a downstream focus. While both nations spend a similar percentage of GDP on total health (combining formal healthcare with “social care” like housing, food, and childcare), their allocation is radically different.

The data suggests the U.S. is not necessarily spending “too much” on health, but that its allocation is radically different from its peers, prioritizing expensive cures over cost-effective prevention.

2. The Three Barriers to Upstream Thinking

Despite the logic of prevention, several powerful forces consistently push individuals and organizations downstream.

A. Problem Blindness: The Invisibility of Solvable Problems

Problem blindness is the belief that negative outcomes are natural, inevitable, or out of one’s control. It is treating a solvable problem like the weather.

• Mechanism: It arises from inattentional blindness (intense focus on one task causing one to miss other information, like radiologists missing a gorilla in a CT scan) and habituation (growing accustomed to consistent stimuli until they become normal).
• Example: Chicago Public Schools (CPS): In 1998, the 52.4% graduation rate was seen by many as an intractable problem caused by poverty and other societal ills—”that’s just how it is.” The problem was accepted as a regrettable but inevitable condition.
• Example: Sexual Harassment: Before the term was coined in 1975 by Lin Farley, the behavior was so normalized that women were often encouraged to tolerate it. Giving the problem a name—”sexual harassment”—was an act of “problematizing the normal,” helping society awaken from problem blindness.
• Example: C-Sections in Brazil: An 84% C-section rate in Brazil’s private health system was seen as normal by many doctors, driven by convenience and financial incentives. An activist movement led by mothers who felt pressured into the procedure successfully challenged this norm, reframing it as a public health problem.

B. Lack of Ownership: “Not My Problem to Fix”

This barrier exists when the people or groups best positioned to solve a problem declare, “That’s not mine to fix.” This can result from fragmented responsibilities, self-interest, or a perceived lack of legitimacy.

• Fragmented Responsibility: At Expedia, no single team was measured on reducing customer calls, so no one “owned” the problem.
• Lack of Psychological Standing: People may feel they lack the legitimacy to act on a problem that doesn’t affect them personally. Research shows that explicitly extending standing (e.g., naming a group “Men and Women Opposed to Proposition 174”) can dramatically increase participation from those without a direct vested interest.
• Taking Ownership: Dr. Bob Sanders & Car Seats: Spurred by a 1975 article in Pediatrics that extended psychological standing to pediatricians on auto safety, Dr. Sanders took ownership of the issue. He successfully lobbied for Tennessee to become the first state to mandate child car seats in 1978. This micro-level action catalyzed a macro-level change, with all 50 states passing similar laws by 1985, saving an estimated 11,274 young lives by 2016.
• Taking Ownership: Ray Anderson & Interface: The founder of carpet-tile firm Interface took ownership of his company’s environmental impact after reading Paul Hawken’s The Ecology of Commerce. He launched “Mission Zero,” a quest to eliminate the company’s negative environmental footprint by 2020. This was an optional, self-imposed burden that transformed the company’s culture and processes.

C. Tunneling: The Tyranny of Short-Term Crises

When experiencing scarcity of time, money, or mental bandwidth, people adopt “tunnel vision.” They stop long-term planning and focus solely on managing the immediate crisis, which prevents upstream thinking.

• The Scarcity Trap: The experience of poverty reduces cognitive capacity more than a full night without sleep. It forces short-sighted decisions (like taking a payday loan) not because people are undisciplined, but because the tunnel of scarcity leaves no room for long-term considerations.
• Organizational Tunneling: A study of nurses found they were constantly engaged in creative workarounds for recurring problems (e.g., missing equipment, lack of towels) but never engaged in fixing the underlying processes. Their scarce time and attention kept them in a reactive mode.
• Escaping the Tunnel: Escaping requires creating slack—a reserve of time or resources dedicated to problem-solving. This can be structured, as with the “safety huddles” in hospitals or the “Freshman Success Teams” at CPS, which provide a guaranteed forum for emerging from the tunnel to address systems-level issues.
• Co-opting the Tunnel: The Ozone Layer: To address the long-term threat of ozone depletion, advocates had to make an upstream problem feel downstream. They co-opted the power of tunneling by creating urgency through public advocacy, the memorable metaphor of an “ozone hole,” and negotiating international agreements like the Montreal Protocol that removed threats for opponents (like DuPont), thus reducing their need to fight the solution.

3. Key Strategies for Upstream Leaders

Successfully navigating the barriers requires addressing a series of fundamental questions.

A. How Will You Unite the Right People?

Upstream work is fundamentally collaborative, requiring leaders to “surround the problem” with all the necessary stakeholders.

• Key Insight: Give every stakeholder a role. Progress hinges on voluntary effort, so maintaining a “big tent” is crucial.
• Case Study: Iceland’s War on Teen Substance Abuse: In the 1990s, 42% of Icelandic teens reported being drunk in the past month. A coalition of researchers, policymakers, schools, parents, and community groups united to change the culture around teens.
  • Strategy: They focused on boosting “protective factors” (e.g., participation in formal sports, time spent with parents, “natural highs”) and reducing “risk factors” (unstructured, unsupervised time).
  • Tactics: They reinforced curfews, gave families “gift cards” for recreational activities, and professionalized coaching in sports clubs.
  • Result: Over 20 years, the percentage of teens getting drunk in the past 30 days fell from 42% to 5%. Daily smoking dropped from 23% to 3%.
• Case Study: Domestic Violence in Newburyport, MA: After a woman was murdered by her estranged husband, the Jeanne Geiger Crisis Center united police, advocates, parole officers, and prosecutors to form a Domestic Violence High Risk Team.
  • Data-Driven Collaboration: The team meets monthly to review cases of women identified by the “Danger Assessment” tool as being at extreme risk of homicide. They use a by-name list to coordinate actions like police drive-bys and creating emergency plans.
  • Result: In the 14 years since the team’s formation, not one woman in the communities they serve has been killed in a domestic violence–related homicide, compared to 8 in the 10 years prior.
• The Role of Data: In many successful upstream efforts, data is not used for top-down “inspection” but for frontline “learning.” Real-time, granular data (like a by-name list) becomes the centerpiece that unites diverse teams around a concrete and shared goal: “What are we going to do about Michael next week?”

B. How Will You Change the System?

Lasting upstream work must culminate in systems change, altering the “water” we swim in so that better outcomes happen by default.

• Systems Determine Probabilities: A well-designed system makes success highly probable (e.g., fluoridated water preventing cavities). A flawed system rigs the game against certain people. As Dr. Anthony Iton discovered, disparities in life expectancy of up to 20 years between nearby ZIP codes are not caused by a few factors, but by entire systems (housing, education, crime, food access) that create “incubators of chronic stress.”
• The California Endowment’s BHC Initiative: This $1 billion, 10-year program aims to fix these broken systems not by directly providing health services, but by empowering residents of 14 challenged communities to gain political power and win policy victories that reshape their environments.
• The Danger of Enabling Bad Systems: Some well-intentioned downstream efforts can inadvertently prop up the flawed systems that create need. For example, while DonorsChoose provides vital classroom supplies, its success could excuse school districts from their funding obligations. The goal should be to push for a world where such crutches are no longer needed.

C. Where Can You Find a Point of Leverage?

In complex systems, the challenge is finding the right lever. This requires getting “proximate” to the problem.

• Case Study: The UChicago Crime Lab & “Becoming a Man” (BAM): To understand youth violence, researchers read 200 consecutive homicide reports. They discovered that many deaths resulted not from strategic gang wars but from impulsive reactions to trivial disputes. This pointed to impulsivity as a leverage point.
  • The Intervention: They funded and studied “Becoming a Man” (BAM), a program that used small-group sessions and cognitive behavioral therapy (CBT) to help at-risk young men learn to manage anger and slow down their thinking in fraught situations.
  • The Result: A randomized controlled trial found that BAM participants had 45% fewer violent-crime arrests.
• The Power of Proximity: Architects designing for the elderly donned an “age simulation suit” to experience navigation challenges firsthand. This direct experience revealed leverage points like the need for more benches, handrails, and three-step escalators.

D. How Will You Get Early Warning of the Problem?

Early warning signals provide the time and maneuvering room to prevent a problem or blunt its impact.

• Predictive Analytics:
  • LinkedIn: Discovered that a customer’s product usage in the first 30 days could predict their likelihood of churning a year later. They shifted resources to intensive onboarding to ensure early engagement.
  • Northwell Health EMS: Uses historical data on 911 calls to predict where emergencies will occur (e.g., near nursing homes at mealtimes) and forward-deploys ambulances to reduce response times.
• Human Sensors:
  • Sandy Hook Promise: After the 2012 school shooting, the organization realized that in most mass shootings, the perpetrator tells someone their plans in advance. They created the “Know the Signs” program to train students to spot warning signs and the “Say Something” anonymous tip line to report them. This system has averted multiple credible school shooting threats and led to hundreds of suicide interventions.
• The Danger of False Positives: Early warning systems can backfire. An “epidemic” of thyroid cancer in South Korea was revealed to be an epidemic of overdiagnosis. Mass screening found huge numbers of slow-growing, nonlethal cancers (“turtles”), leading to unnecessary and harmful treatments for a problem that didn’t exist.

E. How Will You Measure Success and Avoid “Ghost Victories”?

Success in upstream work is often the absence of a negative event, making it hard to measure. This reliance on proxy measures can lead to “ghost victories”—superficial successes that cloak underlying failure.

1. Mistaking Macro Trends for Success: In the 1990s, police chiefs across the U.S. claimed credit for falling crime rates, when in fact they were mostly benefiting from a nationwide trend.
2. Misalignment of Measures and Mission: The City of Boston’s Public Works department measured its sidewalk repair success by spending per zone and 311 cases closed. This led them to fix sidewalks in wealthy neighborhoods (whose residents called 311) while neglecting crumbling sidewalks in poor neighborhoods, undermining their mission of equity and walkability.
3. Measures Becoming the Mission: This is the most destructive form, where people “game” the metrics. The NYPD’s CompStat system, which held precinct leaders accountable for crime statistics, led to the widespread downgrading of crimes. In a chilling example, a reported rape of a prostitute was nearly reclassified as a “theft of service” to keep the numbers down.

To avoid ghost victories, leaders should use paired measures (balancing quantity with quality, as CPS did with graduation rates and ACT scores) and “pre-game” how measures could be misused.

F. How Will You Avoid Doing Harm?

Upstream interventions tinker with complex systems and can create unintended negative consequences, known as the “cobra effect.”

• Case Study: Macquarie Island: A decades-long effort to eradicate invasive species on a subantarctic island created a cascade of problems. Killing rabbits (to stop erosion) led cats to eat rare birds. Killing the cats led to a rabbit population explosion. Killing all pests led to invasive weeds running rampant.
• Anticipating Second-Order Effects: Wise interventions require seeing the whole system. The “cobra effect” is when an attempted solution makes the problem worse. Examples include an open-office plan meant to increase face-to-face collaboration actually causing it to plunge by 70%, or a ban on thin plastic bags leading retailers to offer thicker plastic bags.
• The Need for Feedback Loops: Because not all consequences can be foreseen, upstream work requires experimentation and fast, reliable feedback loops. A business that creates a feedback loop for its staff meetings (rating each meeting on a 1-5 scale) can continuously improve them, whereas most meetings never get better because there is no mechanism for learning.

G. Who Will Pay for What Does Not Happen?

Funding prevention is notoriously difficult because success is invisible and payment models are designed for reaction.

• The “Wrong Pocket Problem”: This occurs when the entity that pays for an intervention is not the one that reaps the financial benefits.
• Case Study: The Nurse-Family Partnership (NFP): This program, which provides nurse home visits to first-time, low-income mothers, has been proven by multiple RCTs to produce significant long-term social benefits (e.g., reduced child abuse, preterm births, crime, and welfare payments), yielding a return of over $6 for every $1 invested. However, it struggles to get funding because the benefits are scattered across many “pockets” (Medicaid, criminal justice, social services), while a single entity is asked to bear the upfront cost.
• Innovative Funding Models:
  • Pay for Success: A model being used in South Carolina to fund NFP, where private investors and foundations provide upfront capital. If the program meets pre-agreed success metrics, the government repays the investors. This shifts the financial risk away from the government.
  • Accountable Care Organizations (ACOs): A model where Medicare shares savings with groups of doctors who succeed in keeping their patients healthier and out of the hospital, creating a direct financial incentive for prevention.

4. Addressing Distant and Improbable Threats (“Far Upstream”)

Upstream thinking can also be applied to one-off, improbable, or unpreventable threats.

• The Prophet’s Dilemma: This is a prediction that prevents what it predicts from happening. The massive global effort to fix the Y2K bug is a prime example. When disaster didn’t strike, many claimed it was a hoax, but it is likely the frantic preparations were what prevented the catastrophe.
• The Power of Rehearsal: The “Hurricane Pam” simulation, conducted 13 months before Hurricane Katrina, convened 300 stakeholders to game-plan a response to a catastrophic New Orleans hurricane. While the eventual Katrina response was a national failure in many respects, the planning from Pam led to a drastically improved “contraflow” evacuation plan, which is credited with reducing the death toll from a projected 60,000 to approximately 1,700. The lesson is that preparing for disaster requires practice, but organizations in a state of “tunneling” often fail to invest in it.
• Existential Risk & The “Black Ball” Hypothesis: Philosopher Nick Bostrom posits that technological invention is like pulling balls from an urn. So far we have pulled white (beneficial) and gray (mixed-blessing) balls. But what if there is a black ball—a technology that is easily accessible and allows a small group to cause mass destruction, thereby destroying civilization? The response to the remote threat of “Moon germs” in the 1960s, which led to the creation of NASA’s Planetary Protection Officer and strict quarantine protocols, provides an early model for how humanity can collectively address improbable but high-stakes risks.

5. Conclusion: You, Upstream

The principles of upstream thinking can be applied by individuals to solve personal and organizational problems.

• Personal Application: Identify recurring problems in life—from finding parking to marital friction—and devise systems to prevent them. The creation of “Daddy Dolls” by a military spouse to ease her children’s pain during deployment is a powerful example of an individual creating an upstream solution.
• Engaging in Societal Problems: When seeking to contribute to larger issues, one should:
  1. Be impatient for action but patient for outcomes: Upstream work is a long game of chipping away at a problem.
  2. Recognize that macro starts with micro: You cannot help a thousand people until you understand how to help one. Deep, proximate understanding is key.
  3. Favor “Scoreboards” over “Pills”: Prioritize initiatives that use real-time data for continuous learning and adaptation (a scoreboard) over those that seek a single, perfect, scalable solution that cannot be changed (a pill).
• The Power of One Person: A single, retiring actuary at the Centers for Medicare & Medicaid Services wrote a “cry of the heart” letter to his boss, successfully arguing that the agency should not count “longer lives” as a cost when evaluating preventive programs. This quiet act of defiance changed a federal rule, unlocking funding for life-saving programs and demonstrating that even within vast bureaucracies, one person can achieve a profound upstream victory.

Upstream Thinking Study Guide

Quiz: Short-Answer Questions

Instructions: Answer the following questions in two to three sentences, drawing exclusively from the information provided in the source context.

1. Describe the public health parable that opens the text. What is the core lesson it is meant to illustrate?
2. Explain the problem Ryan O’Neill discovered at Expedia in 2012. What was the upstream solution the company implemented?
3. What is “problem blindness”? How did this barrier manifest within the Chicago Public Schools (CPS) system regarding its low graduation rate?
4. Define the barrier of “lack of ownership” and the related concept of “psychological standing.” How did the advocates for child car seat laws in the 1970s overcome this barrier?
5. What is “tunneling”? How does this phenomenon, as described by Eldar Shafir and Sendhil Mullainathan, act as a barrier to upstream thinking?
6. Summarize the core philosophy of the “Drug-free Iceland” campaign. What were the “risk factors” and “protective factors” it aimed to influence?
7. What is a “ghost victory”? Using the example of Boston’s sidewalk repairs, explain how an organization can succeed on its metrics while failing its mission.
8. How did the University of Chicago Crime Lab identify “impulsivity” as a key leverage point for reducing youth violence? Describe the “Becoming a Man” (BAM) program that addressed this.
9. Explain the “cobra effect,” using the example of the British administrator’s attempt to reduce the cobra population in Delhi.
10. What is the “wrong pocket problem”? How does the case of the Nurse-Family Partnership (NFP) illustrate this challenge in funding preventive programs?

Essay Questions

Instructions: The following questions are designed to provoke deeper thought and synthesis of the concepts presented in the text. Formulate a detailed response for each, citing specific examples and arguments from the source material.

1. The text identifies three primary barriers to upstream thinking: Problem Blindness, Lack of Ownership, and Tunneling. Analyze how these three barriers were present in the Expedia case study and how the company’s leaders ultimately overcame them to implement a successful upstream intervention.
2. Discuss the role of data in enabling upstream work, contrasting “data for the purpose of learning” with “data for the purpose of inspection.” Use the examples of the Chicago Public Schools’ Freshman On-Track metric, the Newburyport Domestic Violence High Risk Team’s Danger Assessment, and the Rockford homelessness team’s “by-name list” to illustrate your points.
3. Compare and contrast the challenges of upstream interventions in the public sector versus the private sector, using the stories of Ray Anderson at Interface and Dr. Bob Sanders’s campaign for child car seats in Tennessee. What unique advantages and disadvantages did each leader face in trying to solve a problem they chose to own?
4. Upstream interventions often create unintended consequences. Using the case studies of the Macquarie Island pest eradication program and the attempts to ban single-use plastic bags, discuss the importance of systems thinking, experimentation, and feedback loops in avoiding harm.
5. The author argues that our society’s attention is “grossly asymmetrical” and skewed toward downstream reaction rather than upstream prevention. Using the detailed comparison between the United States and Norwegian healthcare systems, analyze the author’s argument. What are the demonstrated benefits and disadvantages of each country’s approach to “buying health”?

Quiz Answer Key

1. The parable describes two friends rescuing drowning children from a river. While one continues the downstream work of pulling kids from the water, the other goes upstream to “tackle the guy who’s throwing all these kids in the water.” The lesson illustrates the difference between reacting to problems (downstream) and preventing them at their source (upstream).
2. Ryan O’Neill found that for every 100 Expedia customers, 58 placed a call for help, with the number one reason being a request for their itinerary. The upstream solution was to prevent these calls by adding an automated voice-response option, improving email delivery to avoid spam filters, and creating an online tool for customers to retrieve their own itineraries.
3. “Problem blindness” is the belief that negative outcomes are natural, inevitable, or out of one’s control. Within CPS, many staff members had come to accept the 50% dropout rate as “just how it is,” believing it was caused by intractable root causes like poverty or lack of student effort, which reinforced a sense of helplessness.
4. “Lack of ownership” means that the parties capable of addressing a problem believe “that’s not mine to fix.” “Psychological standing” is the sense of legitimacy one feels in protesting or acting on an issue. Annemarie Shelness and Seymour Charles overcame this by publishing an article in Pediatrics, extending psychological standing to pediatricians and framing auto safety as a form of preventive medicine for them to own.
5. “Tunneling” is a state of mind caused by scarcity of time, money, or bandwidth, where people adopt a narrow, short-term focus on immediate problems. It is a barrier to upstream thinking because it confines people to reactive problem-solving and prevents them from engaging in the long-term planning and systems thinking required to prevent future problems.
6. The core philosophy was to change the community and cultural environment surrounding teenagers to make substance use feel abnormal. The campaign worked to reduce risk factors, such as unstructured time and friends who drink, while boosting protective factors, like participation in formal sports and spending more time with parents.
7. A “ghost victory” is a superficial success that cloaks an underlying failure, often occurring when short-term measures do not align with the long-term mission. Boston’s Public Works department succeeded on its measures of closing 311 cases and spending its budget, but this system disproportionately repaired sidewalks in wealthy neighborhoods, failing the ultimate mission of equity and walkability for all citizens.
8. By studying 200 homicide reports, the Crime Lab found that many deaths resulted not from strategic gang activity but from impulsive reactions to trivial disputes, like arguments over a bike or a basketball game. The “Becoming a Man” (BAM) program used cognitive behavioral therapy (CBT) and group mentoring to teach young men to slow down their thinking and manage anger in fraught situations.
9. The “cobra effect” occurs when an attempted solution makes the problem worse. In colonial Delhi, a British administrator offered a bounty for dead cobras to reduce their population. In response, citizens began farming cobras to collect the bounty, and when the program was canceled, they released their now-worthless snakes, resulting in more cobras than before.
10. The “wrong pocket problem” occurs when the entity that pays for a preventive intervention does not receive the primary financial benefit from its success. The Nurse-Family Partnership has been proven to save society money by reducing crime, preterm births, and welfare payments, but it struggles to get funding because these savings are scattered across many different government “pockets” (criminal justice, Medicaid, etc.), none of which want to bear the full upfront cost.

Glossary of Key Terms

Contact Factoring Specialist, Chris Lehnes