Modern institutions increasingly understand the world through numbers. Performance is translated into percentages, risk into scores, progress into indicators, efficiency into ratios, and social conditions into rankings. Dashboards display whether targets have been achieved. Key performance indicators show which units are performing well. Indexes compare cities, regions, organizations, schools, hospitals, and even countries. Numbers appear to offer something institutions have always wanted: a clear way to transform complex realities into information that can be compared, monitored, and acted upon.
There are good reasons for this attraction. Institutions cannot govern large societies through intuition alone. Public administrations need statistics to allocate resources. Organizations need indicators to identify problems. Governments need measurable information to evaluate policies. Without numbers, many forms of coordination would become slower, less consistent, and more vulnerable to arbitrary judgment.
The problem therefore does not begin with measurement itself.
It begins when measurement slowly becomes the dominant language through which institutions understand reality.
A number can describe part of the world with extraordinary precision, while still failing to capture what matters most about it. A percentage can show that a target has been achieved without revealing how that achievement was produced. A ranking can distinguish high performers from low performers while hiding differences in context. A dashboard can signal that everything is green even when important problems remain outside the indicators being monitored.
Numbers help institutions see.
But every way of seeing also creates a way of not seeing.
The Appeal of Quantification
Quantification is powerful because it simplifies complexity. A complicated social condition can be translated into a metric that can be compared across locations and time. Once translated into numbers, situations that initially appear different can be placed within a common frame.
Wendy Nelson Espeland and Mitchell Stevens describe this process as commensuration: the transformation of different qualities into a common metric that allows comparison. Their work emphasizes that commensuration is not merely a technical activity. It is also a social process with cognitive and political consequences because choosing what can be compared requires decisions about which differences matter and which differences can be ignored.
This capacity is indispensable to modern administration. A government managing millions of citizens cannot rely exclusively on individual narratives. It needs categories. It needs standardized records. It needs statistics that allow officials to recognize patterns that would otherwise remain invisible.
Numbers also appear to reduce disagreement. Two people may have different interpretations of whether an institution is performing well, but an indicator seems to provide a neutral point of reference. A target of 90 percent appears clearer than the statement that performance has “improved significantly.”
The apparent objectivity of numbers is therefore deeply attractive. Quantification gives administrative decisions a language that seems detached from personal preference.
Yet numbers do not enter institutions without prior choices.
Someone decides what should be measured.
Someone defines the unit.
Someone determines the threshold.
Someone decides which variables should be included and which should remain outside the calculation.
The resulting number may be objective within its methodology, but the methodology itself reflects assumptions about what counts as relevant.
From Measurement to Governance
Measurement becomes more consequential when indicators stop merely describing institutional performance and begin directing it.
A metric originally designed to observe behavior can eventually influence that behavior. Once an indicator becomes connected to evaluation, funding, promotion, reputation, sanctions, or institutional legitimacy, people begin organizing their activities around the indicator itself.
The difference may appear subtle, but it changes the function of measurement.
At first, numbers describe institutions.
Eventually, institutions begin responding to numbers.
A public office measured by the speed of completing cases will naturally devote attention to speed. A university evaluated through publication counts will encourage publication. A hospital assessed through selected treatment indicators will focus organizational energy on those indicators. A company evaluated by quarterly growth will become sensitive to activities that influence quarterly results.
This does not mean that such indicators are inappropriate. Measurement can encourage accountability and reveal poor performance that would otherwise remain hidden.
The difficulty emerges when achieving the indicator becomes indistinguishable from achieving the underlying purpose.
Theodore Porter, in Trust in Numbers, examined the historical and institutional attraction of quantitative objectivity, particularly in environments where organizations seek forms of decision-making that appear impersonal and defensible. Numbers can create trust precisely because they seem to limit discretion. Yet this strength can also encourage institutions to treat quantified measures as substitutes for judgment rather than as instruments that support it.
Modern governance increasingly operates within this tension.
Institutions need numbers to make decisions.
But decisions cannot always be reduced to numbers.
Indicators Do More Than Measure
An indicator does not merely represent reality. It can reorganize attention.
Once something is measured systematically, it becomes easier to see inside an institution. It appears in meetings, performance reports, presentations, dashboards, and evaluations. Issues without equivalent indicators often struggle to compete for attention.
Measurement therefore creates visibility.
This visibility produces institutional consequences.
Imagine two problems within the same organization. The first has a clearly defined indicator that is updated every week and displayed on a dashboard. The second is difficult to quantify and depends upon qualitative observation. Even if both are equally important, the first problem is likely to receive more systematic attention because its existence can be demonstrated repeatedly through data.
Over time, organizations may become particularly responsive to what they can measure.
This creates a subtle inversion. Indicators are initially developed to help institutions understand reality, but institutional routines can gradually encourage reality to be understood primarily through indicators.
Sally Engle Merry examined this problem in her work on global indicators. She argued that indicators translate complex social phenomena into simplified quantitative forms that can guide organizations and governments. Yet this translation necessarily removes some context and meaning. The resulting measure may become extremely influential even though it represents only one constructed interpretation of the phenomenon being measured.
The challenge is therefore not whether indicators are true or false in a simple sense.
The deeper question is what kind of reality becomes visible through them.
The Politics Hidden Inside Metrics
Numbers often appear politically neutral because they are expressed mathematically.
But the construction of metrics frequently contains normative choices.
Consider an index designed to measure development. Should development be defined primarily through income, infrastructure, education, health, access to public services, environmental quality, or social equality? Different choices produce different representations of the same society.
The mathematics may be rigorous after the variables have been selected.
Yet variable selection is not a mathematical decision alone.
The same applies inside organizations. A performance system that prioritizes speed implicitly communicates that speed matters. A metric emphasizing volume communicates that output matters. An indicator emphasizing satisfaction communicates that user perception matters.
Every system of measurement contains an institutional theory about what deserves attention.
Sometimes that theory is explicit.
Often it remains hidden inside the indicator.
This is why debates about measurement are ultimately debates about values. The issue is not simply whether the calculation is correct, but whether the calculation represents the right thing.
Quantification can make political choices appear technical.
A discussion that might otherwise involve competing values can become a debate over methodologies, formulas, thresholds, and datasets. Technical language can make those choices appear more neutral than they actually are.
The politics has not disappeared.
It has moved into the architecture of measurement.
The Dashboard as Institutional Vision
The dashboard has become one of the defining administrative instruments of digital organizations.
Its attraction is obvious. Instead of reading hundreds of pages of reports, leaders can observe the organization through a small number of indicators displayed visually. Red identifies problems. Green signals acceptable performance. Trends reveal whether conditions are improving or deteriorating.
Dashboards can dramatically improve situational awareness.
But they also create a particular form of institutional vision.
A dashboard does not show an organization.
It shows a representation of an organization based on what the system has been designed to measure.
This distinction matters.
An institution may contain thousands of experiences, interactions, conflicts, informal practices, exceptional circumstances, and contextual variations. Only a fraction of them appear on the dashboard.
The dashboard therefore reduces organizational complexity into a manageable representation.
Such reduction is necessary.
No decision-maker can absorb everything.
The danger arises only when the representation is mistaken for the institution itself.
A green indicator can create confidence even when important problems remain unmeasured. A red indicator can create urgency even when the underlying situation requires contextual interpretation.
Decision-makers therefore need a form of literacy that goes beyond reading the numbers.
They must also understand the boundaries of the numbers.
Performance and the Logic of Targets
Targets can produce extraordinary organizational energy.
A clearly defined objective creates direction. Teams know what is expected. Leaders can evaluate progress. Resources can be concentrated around measurable goals.
But targets can also reshape behavior in unexpected ways.
Once performance becomes closely connected to numerical objectives, individuals and organizations have incentives to improve the metric. Sometimes improving the metric and improving reality are the same thing.
Sometimes they are not.
A system evaluated primarily through quantity may increase production while reducing quality. A service evaluated primarily through completion time may process cases faster while spending less time on complicated cases. An institution rewarded for reducing unresolved cases may develop practices that improve administrative statistics without addressing underlying problems.
The phenomenon is often associated with Goodhart’s Law, commonly summarized as the idea that when a measure becomes a target, it can cease to function as a good measure.
The principle is important because it reveals the behavioral dimension of measurement.
People do not merely exist inside systems of indicators.
They learn how those systems work.
Once incentives become attached to metrics, rational actors adjust their behavior accordingly.
This is not necessarily manipulation. Often it is simply adaptation.
Organizations respond to what organizations are asked to achieve.
The deeper problem therefore lies in designing measurement systems that preserve the relationship between institutional purpose and measurable performance.
Governance by Numbers
Alain Supiot describes a broader transformation in which governance increasingly revolves around measurable objectives. In Governance by Numbers, he argues that contemporary governance has progressively expanded the normative role of quantification, with objectives and performance mechanisms increasingly shaping the operation of institutions.
This does not mean that law, political authority, or administrative judgment have disappeared.
Rather, another mode of governance has become more influential alongside them.
Instead of asking only whether rules have been followed, institutions increasingly ask whether targets have been achieved.
Instead of evaluating only legality, they evaluate performance.
Instead of relying exclusively on hierarchical commands, they use benchmarks, indicators, rankings, and comparative assessments.
The transformation can produce more adaptive institutions. It may also create organizations that continuously measure themselves.
This continuous measurement can change institutional culture.
Employees become increasingly aware of how their actions appear within measurement systems. Managers organize priorities around metrics. Strategic conversations begin with dashboards. Performance becomes visible through charts.
Institutional life becomes translated into measurable signals.
The question is whether everything worth governing can be translated in this way.
The Things Numbers Cannot Easily See
Many of the most important dimensions of institutional life resist simple measurement.
Trust is difficult to quantify.
Professional judgment is difficult to standardize.
Fairness often depends upon context.
Institutional resilience may remain invisible until disruption occurs.
Human dignity cannot easily be reduced to an indicator.
The quality of a decision is not always captured by the speed with which it was made.
The significance of these dimensions does not make measurement impossible. Researchers continually develop sophisticated methods to measure complex social phenomena.
The issue is different.
Every measurement necessarily captures some dimensions more effectively than others.
A number can therefore be simultaneously useful and incomplete.
This is particularly important in public institutions because government frequently deals with exceptional cases. Standardization allows equal treatment, but circumstances sometimes require judgment. Indicators allow large-scale management, but individual situations may not fit neatly into statistical categories.
The ability to recognize exceptions is part of institutional intelligence.
A system governed exclusively through metrics risks confusing consistency with justice.
Numbers and the Distribution of Attention
Perhaps the most significant power of numbers is not their ability to determine decisions directly.
It is their ability to determine what institutions notice.
Issues supported by abundant data become easier to discuss. Problems that can be visualized become easier to communicate. Performance that can be compared becomes easier to evaluate.
Meanwhile, phenomena that are difficult to quantify may remain institutionally quiet.
This produces what might be called an inequality of measurability.
Some realities generate continuous streams of data.
Others leave few traces.
Digital systems intensify this asymmetry because organizations increasingly depend on information that can enter databases. What cannot be recorded may become difficult to analyze, and what cannot be analyzed may struggle to influence decisions.
The result can be paradoxical.
Institutions may possess more data than at any previous moment while still failing to understand important parts of the societies they govern.
More information does not automatically produce deeper knowledge.
The difference lies in interpretation.
Data Is Not Understanding
Contemporary organizations sometimes speak as if enough data will eventually eliminate uncertainty.
The assumption is understandable. Better information usually improves decisions. Larger datasets can reveal patterns that smaller datasets cannot detect.
Yet uncertainty is not produced only by insufficient information.
Some uncertainty exists because social reality itself is complex.
People change their behavior.
Institutions interact.
Policies produce unintended effects.
Economic conditions shift.
Cultural meanings evolve.
Events occur that historical data cannot anticipate.
Data can reduce uncertainty, but it cannot eliminate the need for interpretation.
This is where human judgment remains essential.
Judgment connects numbers with context.
It asks whether an indicator accurately represents current conditions. It recognizes exceptional circumstances. It distinguishes correlation from meaningful explanation. It considers whether a technically efficient decision is institutionally appropriate.
Numbers can reveal patterns.
Understanding requires something more.
Governing With Numbers
The alternative to governing by numbers is not governing without numbers.
Modern societies cannot abandon statistics, indicators, databases, or performance measurement. Doing so would weaken accountability and make many forms of public administration impossible.
The more productive distinction is between governing by numbers and governing with numbers.
Governing by numbers allows metrics to become the dominant logic of decision-making.
Governing with numbers treats quantitative information as one component of institutional judgment.
The difference is significant.
In the first approach, indicators increasingly define reality.
In the second, indicators help institutions investigate reality.
Numbers should provoke questions rather than eliminate them.
Why is this indicator improving?
Why is another declining?
What conditions produced the change?
Are there experiences that the data does not capture?
Do the metrics remain aligned with the institution’s original purpose?
These questions restore interpretation to the center of governance.
They also recognize that accountability involves more than achieving targets.
It involves understanding consequences.
Beyond the Measurable Institution
The rise of data-driven governance represents one of the most important transformations of contemporary institutional life.
Public agencies, companies, universities, hospitals, platforms, and international organizations increasingly operate through systems of measurement. The trend is unlikely to reverse. Artificial intelligence, predictive analytics, real-time dashboards, and automated monitoring will make quantification even more central to organizational decision-making.
The challenge will therefore not be resisting numbers.
It will be learning how to live intelligently with them.
Institutions must remain capable of distinguishing between measurement and meaning, between indicators and reality, between performance and purpose.
A society governed through data needs institutions that understand both the power and the limits of quantification.
Because numbers are never merely numbers once they enter systems of authority. They begin influencing attention, incentives, priorities, and eventually behavior.
The most important question is therefore not whether institutions should measure more or less.
It is whether they still remember why they began measuring in the first place.
Numbers are powerful because they make reality manageable.
Wisdom begins by remembering that reality is always larger than the number used to describe it.
References
Espeland, W. N., & Stevens, M. L. (1998). Commensuration as a social process. Annual Review of Sociology, 24, 313–343. https://doi.org/10.1146/annurev.soc.24.1.313
Merry, S. E. (2016). The Seductions of Quantification: Measuring Human Rights, Gender Violence, and Sex Trafficking. University of Chicago Press.
Porter, T. M. (1995). Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton University Press.
Supiot, A. (2017). Governance by Numbers: The Making of a Legal Model of Allegiance. Hart Publishing.
Zuboff, S. (2019). The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs.
