Technochauvinism


/ˈtɛknoʊˈʃoʊvəˌnɪzəm/

The belief that technology is always the solution (Broussard, 2018).

Since the beginning of its rise in the late 20th century, digital and computer technology promised to improve many ways in which society operates. Personal computers, mobile phones, and the internet are some of the most ubiquitous examples of technology that have demonstrable capabilities to make lives easier to a certain extent.

However, recent years have shown increasing techlash—defined by The Oxford English Dictionary as “a strong and widespread negative reaction to the growing power and influence of large technology companies, particularly those based in Silicon Valley”—as a response to the harm that technology has helped create. Misinformation, privacy violation, and algorithmic bias are phrases that can often be found in the same sentence as one or more tech companies.

Computer scientist and data journalist Meredith Boussard, who is a professor at New York University, argues that these problems stem from technochauvinism—the belief that technology is always the solution. The summary of her book, Artificial Unintelligence, writes:

“… it’s just not true that social problems would inevitably retreat before a digitally enabled Utopia. To prove her point, she undertakes a series of adventures in computer programming. She goes for an alarming ride in a driverless car, concluding “the cyborg future is not coming any time soon”; uses artificial intelligence to investigate why students can’t pass standardized tests; deploys machine learning to predict which passengers survived the Titanic disaster; and attempts to repair the U.S. campaign finance system by building AI software. If we understand the limits of what we can do with technology, Broussard tells us, we can make better choices about what we should do with it to make the world better for everyone.”

The term technochauvinism is similar to technosolutionism. In that, they both describe the belief that most, if not all, complex issues can be solved with the right computation and engineering. However, the use of “chauvinism” is intentional because part of the criticism is about the rampant gender inequality in the tech industry, which manifest in many ways including algorithmic sexism.

“In Artificial Unintelligence, Meredith Broussard argues that our collective enthusiasm for applying computer technology to every aspect of life has resulted in a tremendous amount of poorly designed systems. We are so eager to do everything digitally—hiring, driving, paying bills, even choosing romantic partners—that we have stopped demanding that our technology actually work. Broussard, a software developer and journalist, reminds us that there are fundamental limits to what we can (and should) do with technology. With this book, she offers a guide to understanding the inner workings and outer limits of technology—and issues a warning that we should never assume that computers always get things right.”

Nowcasting


/naʊˈkæstɪŋ/

A method of describing the present or the near future by analyzing datasets that are not traditionally included in the analysis (e.g. web searches, reviews, social media data, etc.)

Nowcasting is a term that originates in meteorology, which refers to “the detailed description of the current weather along with forecasts obtained by extrapolation for a period of 0 to 6 hours ahead.” Today, nowcasting is also used in other fields, such as macroeconomics and health, to provide more up-to-date statistics.

Traditionally, macroeconomic statistics are collected on a quarterly basis and released with a substantial lag. For example, GDP data for the euro area “is only available at quarterly frequency and is released six weeks after the close of the quarter.” Further, economic datasets from government agencies such as the US Census Bureau “typically appear only after multi-year lags, and the public-facing versions are aggregated to the county or ZIP code level.

The arrival of the big data era has shown some promise to improve nowcasting. A paper by Edward L. Glaeser, Hyunjin Kim, and Michael Luca presents “evidence that Yelp data can complement government surveys by measuring economic activity in close to real-time, at a granular level, and at almost any geographic scale.” In the paper, the authors concluded:

“Our analyses of one possible data source, Yelp, suggests that these new data sources can be a useful complement to official government data. Yelp can help predict contemporaneous changes in the local economy. It can also provide a snapshot of economic change at the local level. It is a useful addition to the data tools that local policy-makers can access.

“Yet our analysis also highlights the challenges with the idea of replacing the Census altogether at any point in the near future. Government statistical agencies invest heavily in developing relatively complete coverage, for a wide set of metrics. The variation in coverage inherent in data from online platforms make it difficult to replace the role of providing official statistics that government data sources play.

“Ultimately, data from platforms like Yelp –combined with official government statistics – can provide valuable complementary datasets that will ultimately allow for more timely and granular forecasts and policy analyses, with a wider set of variables and more complete view of the local economy.”

Another example comes from the United States Federal Reserve (The Fed), which used data from payroll-processing company ADP to payroll employment. This data is traditionally provided by Current Employment Statistics (CES) survey. Despite being “one of the most carefully conducted measures of labor market activity and uses an extremely large sample, it is still subject to significant sampling error and nonsampling errors.” The Fed sought to improve the reliability of this survey by including data provided by ADP. The study found that combining CES and ADP data “reduces the error inherent in both data sources.”

However, nowcasting using big data comes with some limitations. Several researchers evaluated the accuracy of Google Flu Trends (GFT) in the 2012-2013 and 2013-2014 seasons. GFT uses flu-related google searches to make its prediction. The study found that GFT data showed significant overestimation compared to Centers for Disease Control and Prevention (CDC) flu trends prediction.

Jesse Dunietz wrote in Nautilus describing how to address the limitations of big data and make nowcasting efforts more accurate: 

“But when big data isn’t seen as a panacea, it can be transformative. Several groups, like Columbia University researcher Jeffrey Shaman’s, for example, have outperformed the flu predictions of both the CDC and GFT by using the former to compensate for the skew of the latter. “Shaman’s team tested their model against actual flu activity that had already occurred during the season,” according to the CDC. By taking the immediate past into consideration, Shaman and his team fine-tuned their mathematical model to better predict the future. All it takes is for teams to critically assess their assumptions about their data.”

Bilingual


/baɪˈlɪŋgwəl/

Practitioners across disciplines who possess both domain knowledge and data science expertise.

The Governance Lab (GovLab) just launched the 100 Questions Initiative, “an effort to identify the most important societal questions whose answers can be found in data and data science if the power of data collaboratives is harnessed.”

The initiative will seek to identify questions that could help unlock the potential of data and data science in solving various global and domestic issues, including but not limited to, climate change, economic inequality, and migration. These questions will be sourced from individuals who have expertise in both a public issue and data science or what The GovLab calls “bilinguals.”

Tom Kalil, the Chief Innovation Officer at Schmidt Futures, argues that the emergent use of data science and machine learning in the public sector will increase the demand for individuals “who speak data science and social sector.”

Similarly, within the business context, David Meer wrote that “being bilingual isn’t just a matter of native English speakers learning how to conjugate verbs in French or Spanish. Rather, it’s important that businesses cultivate talent that can simultaneously speak the language of advanced data analysis and nuts-and-bolts business operations. As data analysis becomes a more prevalent and powerful lever for strategy and growth, organizations increasingly need bilinguals to form the bridge between the work of advanced data scientists and business decision makers.”

For more info, visit www.the100questions.org

Digital Serfdom


/ˈdɪʤətəl ˈsɜrfdəm/

A condition where consumers give up their personal and private information in order to be able to use a particular product or service.

Serfdom is a system of forced labor that exists in a feudalistic society. It was very common in Europe during the medieval age. In this system, serfs or peasants do a variety of labor for their lords in exchange for protection from bandits and a small piece of land that they can cultivate for themselves. Serfs are also required to pay some form of tax often in the form of chickens or crops yielded from their piece of land.

Hassan Khan in The Next Web points out that the decline of property ownership is indicative that we are living in digital serfdom. In an article he says:

“The percentage of households without a car is increasing. Ride-hailing services have multiplied. Netflix boasts over 188 million subscribers. Spotify gains ten million paid members every five to six months.

“The model of “impermanence” has become the new normal. But there’s still one place where permanence finds its home, with over two billion active monthly users, Facebook has become a platform of record for the connected world. If it’s not on social media, it may as well have never happened.”

Joshua A. T. Fairfield elaborates this phenomenon in his book Owned: Property, Privacy, and the New Digital Serfdom. Fairfield discusses his book in an article in The Conversation, stating that:

“The issue of who gets to control property has a long history. In the feudal system of medieval Europe, the king owned almost everything, and everyone else’s property rights depended on their relationship with the king. Peasants lived on land granted by the king to a local lord, and workers didn’t always even own the tools they used for farming or other trades like carpentry and blacksmithing.

[…]

“Yet the expansion of the internet of things seems to be bringing us back to something like that old feudal model, where people didn’t own the items they used every day. In this 21st-century version, companies are using intellectual property law – intended to protect ideas – to control physical objects consumers think they own.”

In other words, Fairfield is suggesting that the devices and services that we use—iPhones, Fitbits, Roomba, digital door locks, Spotify, Uber, and many more—are constantly capturing data about behaviors. By using these products, consumers have no choice but to trade their personal data in order to access the full functionalities of these devices or services. This data is used by private corporations for targeted advertisement, among others. This system of digital serfdom binds consumers to private corporations that dictate the terms of use for their products or services.

Janet Burns wrote about Alex Rosenblat’s UBERLAND: How Algorithms Are Rewriting The Rules Of Work and gave some examples of how algorithms use personal data to manipulate consumers’ behaviors:

“For example, algorithms in control of assigning and pricing rides have often surprised drivers and riders, quietly taking into account other traffic in the area, regionally adjusted rates, and data on riders and drivers themselves.

“In recent years, we’ve seen similar adjustments happen behind the scenes in online shopping, as UBERLAND points out: major retailers have tweaked what price different customers see for the same item based on where they live, and how feasibly they could visit a brick-and-mortar store for it.”

To conclude, an excerpt from Fairfield’s book cautions: 

“In the coming decade, if we do not take back our ownership rights, the same will be said of our self-driving cars and software-enabled homes. We risk becoming digital peasants, owned by software and advertising companies, not to mention overreaching governments.”

Sources and Further Readings:

Self-Sovereign Identity


/sɛlf-ˈsɑvrən aɪˈdɛntəti/

A decentralized identification mechanism that gives individuals control over what, when, and to whom their personal information is shared.

An identification document (ID) is a crucial part of every individual’s life, in that it is often a prerequisite for accessing a variety of services—ranging from creating a bank account to enrolling children in school to buying alcoholic beverages to signing up for an email account to voting in an election—and also a proof of simply being. This system poses fundamental problems, which a field report by The GovLab on Blockchain and Identity frames as follows:

“One of the central challenges of modern identity is its fragmentation and variation across platform and individuals. There are also issues related to interoperability between different forms of identity, and the fact that different identities confer very different privileges, rights, services or forms of access. The universe of identities is vast and manifold. Every identity in effect poses its own set of challenges and difficulties—and, of course, opportunities.”

A report published in New America echoed this point, by arguing that:

“Societally, we lack a coherent approach to regulating the handling of personal data. Users share and generate far too much data—both personally identifiable information (PII) and metadata, or “data exhaust”—without a way to manage it. Private companies, by storing an increasing amount of PII, are taking on an increasing level of risk. Solution architects are recreating the wheel, instead of flying over the treacherous terrain we have just described.”

SSI is dubbed as the solution for those identity problems mentioned above. Identity Woman, a researcher and advocate for SSI, goes even further by arguing that generating “a digital identity that is not under the control of a corporation, an organization or a government” is essential “in pursuit of social justice, deep democracy, and the development of new economies that share wealth and protect the environment.”

To inform the analysis of blockchain-based Self-Sovereign Identity (SSI), The GovLab report argues that identity is “a process, not a thing” and breaks it into a 5-stage lifecycle, which are provisioning, administration, authentication, authorization, and auditing/monitoring. At each stage, identification serves a unique function and poses different challenges.

With SSI, individuals have full control over how their personal information is shared, who gets access to it, and when. The New America report summarizes the potential of SSI in the following paragraphs:

“We believe that the great potential of SSI is that it can make identity in the digital world function more like identity in the physical world, in which every person has a unique and persistent identity which is represented to others by means of both their physical attributes and a collection of credentials attested to by various external sources of authority.”

[…]

“SSI, in contrast, gives the user a portable, digital credential (like a driver’s license or some other document that proves your age), the authenticity of which can be securely validated via cryptography without the recipient having to check with the authority that issued it. This means that while the credential can be used to access many different sites and services, there is no third-party broker to track the services to which the user is authenticating. Furthermore, cryptographic techniques called “zero-knowledge proofs” (ZKPs) can be used to prove possession of a credential without revealing the credential itself. This makes it possible, for example, for users to prove that they are over the age of 21 without having to share their actual birth dates, which are both sensitive information and irrelevant to a binary, yes-or-no ID transaction.”

Some case studies on the application of SSI in the real world presented on The GovLab Blockchange website include a government-issued self-sovereign ID using blockchain technology in the city of Zug in Switzerland; a mobile election voting platform, secured via smart biometrics, real-time ID verification and the blockchain for irrefutability piloted in West Virginia; and a blockchain-based land and property transaction/registration in Sweden.

Nevertheless, on the hype of this new and emerging technology, the authors write:

“At their core, blockchain technologies offer new capacity for increasing the immutability, integrity, and resilience of information capture and disclosure mechanisms, fostering the potential to address some of the information asymmetries described above. By leveraging a shared and verified database of ledgers stored in a distributed manner, blockchain seeks to redesign information ecosystems in a more transparent, immutable, and trusted manner. Solving information asymmetries may turn out to be the real contribution of blockchain, and this—much more than the current enthusiasm over virtual currencies—is the real reason to assess its potential.

“It is important to emphasize, of course, that blockchain’s potential remains just that for the moment—only potential. Considerable hype surrounds the emerging technology, and much remains to be done and many obstacles to overcome if blockchain is to achieve the enthusiasts’ vision of “radical transparency.”

Further readings:

Grey Data


/greɪ ˈdeɪtə/

Data accumulated by an institution for operational purposes that does not fall under any traditional data protection policies.

Organizations across all sectors accumulate a massive amount of data just by virtue of operating alone, and universities are among such organizations. In a paper, Christine L. Borgman categorizes these as grey data and further suggested that universities should take a lead in demonstrating stewardship of these data, which include student applications, faculty dossier, registrar records, ID card data, security cameras, and many others.

“Some of these data are collected for mandatory reporting obligations such as enrollments, diversity, budgets, grants, and library collections. Many types of data about individuals are collected for operational and design purposes, whether for instruction, libraries, travel, health, or student services.”

(Borgman, p. 380)

Grey data typically does not fall under traditional data protection policies such as Health Insurance Portability and Accountability Act (HIPAA), Family Educational Rights and Privacy Act (FERPA), or Institutional Review Boards. Consequently, there are a lot of debates about how to use (or misuse) them. Borgman points out that universities have been “exploiting these data for research, learning analytics, faculty evaluation, strategic decisions, and other sensitive matters.” On top of this, for-profit companies “are besieging universities with requests for access to data or for partnerships to mine them.”

Recognizing both the value of data and the risks arising from the accumulation of grey data, Borgman proposes a model of Data Stewardship by drawing on the practices of data protection at the University of California which concern information security, data governance, and cyber risk.

This model is an example of a good Data Stewardship practice that the GovLab is advocating amidst the rise of public-private collaboration in leveraging data for public good.

The GovLab’s Data Stewards website presents the need for such practice as follows:

“With these new practices of data collaborations come the need to reimagine roles and responsibilities to steer the process of using private data, and the insights it can generate, to address some of society’s biggest questions and challenges: Data Stewards.

“Today, establishing and sustaining these new collaborative and accountable approaches requires significant and time-consuming effort and investment of resources for both data holders on the supply side, and institutions that represent the demand. By establishing Data Stewardship as a function, recognized within the private sector as a valued responsibility, the practice of Data Collaboratives can become more predictable, scaleable, sustainable and de-risked.”

Sources and Further Readings:

Rawification


/rɑwəfɪˈkeɪʃən/

A process of making datasets raw in three steps: reformatting, cleaning, and ungrounding (Denis and Goeta).

Hundreds of thousands of datasets are now made available via numerous channels from both public and private domains. Based on the stage of processing, these datasets can be categorized as either raw data or processed data. According to an Open Government Data principle, raw data (or primary data) “are published as collected at the source, with the finest possible level of granularity, not in aggregate or modified forms.” While processed data is data that has been through some sort of adulteration, categorization, codification, aggregation, and other similar processes.

A large amount of data that is made publicly available come in processed form. For example, population, trade, and budget data are often presented in aggregated forms, preventing researchers from understanding the underlying stories behind these data, such as the differences in patterns or trends when gender, location, or other variables come into factor. Therefore, a rawification process is oftentimes needed in order for a dataset to be useful for a more detailed, secondary, and valuable analysis.

Jérôme Denis and Samuel Goëta define ‘rawification’ as a process of reformatting, cleaning, and ungrounding data in order to obtain a truly ‘raw’ datasets.

According to Denis and Goëta, reformatting data means making sure that data that has been opened can also be easily readable by the users. This is usually achieved by reformatting the data so that it can be read and manipulated by most processing programs. One of the most commonly used formats is CSV (Comma Separated Values).

The next step in a rawification process is cleaning. In this stage, cleaning means correcting mistakes within the datasets, which include but are not limited to, redundancies and incoherence. In many cases, datasets can have multiple entries for the same item, for example ‘New York University’ and ‘NYU’ might be interpreted as two different entities, or ‘the GovLab’ and ‘the Governance Lab’ might experience a similar issue. Cleaning helps address issues like this.

The final step in a rawification process is ungrounding, which means taking out any ties or links from previous data use. Such ties include color coding, comments, and subcategories. This way the datasets can be purely raw and free of all associations and bias.

Opening up data is a clear step for increasing public access to information held within institutions. However, in order to ensure the utility of that data for those accessing it, a rawification process will likely be necessary.

Additional resources:

Data Fiduciary


/ˈdeɪtə fəˈduʃiˌɛri/

A person or a business that manages individual data in a trustworthy manner. Also ‘information fiduciary’, ‘data trust’, or ‘data steward’.

‘Fiduciary’ is an old concept in the legal world. Its Latin origin is fidere, which means to trust. In the legal context, a fiduciary is usually a person that is trusted to make a decision on how to manage an asset or information, within constraints given by another person who owns such asset or information. Examples of a fiduciary relationship include homeowner and property manager, patient and doctor, or client and attorney. The latter has the ability to make decisions about the trusted asset that falls within the conditions agreed upon by the former.

Jack M. Balkin and Jonathan Zittrain wrote a case for “information fiduciary”, in which they pointed out the urgency of adopting the practice of fiduciary in the data space. In The Atlantic, they wrote:

“The information age has created new kinds of entities that have many of the trappings of fiduciaries—huge online businesses, like Facebook, Google, and Uber, that collect, analyze, and use our personal information—sometimes in our interests and sometimes not. Like older fiduciaries, these businesses have become virtually indispensable. Like older fiduciaries, these companies collect a lot of personal information that could be used to our detriment. And like older fiduciaries, these businesses enjoy a much greater ability to monitor our activities than we have to monitor theirs. As a result, many people who need these services often shrug their shoulders and decide to trust them. But the important question is whether these businesses, like older fiduciaries, have legal obligations to be trustworthy. The answer is that they should.”

Recent controversy involving Facebook data and Cambridge Analytica provides another reason for why companies collecting data from users need to act as a fiduciary. Within this framework, individuals would have a say over how and where their data can be used.

Another call for a form of data fiduciary comes from Google’s Sidewalk Labs project in Canada. After collecting data to inform urban planning in the Quayside area in Toronto, Sidewalk Labs announced that they would not be claiming ownership over the data that they collected and that the data should be “under the control of an independent Civic Data Trust.”

In a blog post, Sidewalk Labs wrote that:

“Sidewalk Labs believes an independent Civic Data Trust should become the steward of urban data collected in the physical environment. This Trust would approve and control the collection of, and manage access to, urban data originating in Quayside. The Civic Data Trust would be guided by a charter ensuring that urban data is collected and used in a way that is beneficial to the community, protects privacy, and spurs innovation and investment.”

Realizing the potential of creating new public value through an exchange of data, or data collaboratives, the GovLab “ is advancing the concept and practice of Data Stewardship to promote responsible data leadership that can address the challenges of the 21st century.” A Data Steward mirrors some of the responsibilities of a data fiduciary, in that they are “responsible for determining what, when, how and with whom to share private data for public good.”

Balkin and Zittrain suggest that there is an asymmetrical power between companies that collect user-generated data and the users themselves, in that these companies are becoming indispensable and having more control over an individual’s data. However, these companies are currently not legally obligated to be trustworthy, meaning that there is no legal consequence for when they use this data in a way that breaches privacy or is in the least interest of the customers.

Under a data fiduciary framework, individuals who are trusted with data are attached with legal rights and responsibilities regarding the use of the data. In a case where a breach of trust happens, the trustee will have to face legal consequences.

Sources and Further Readings:

Commonism


/ˈkɑmənɪz(ə)m/

“A new radical, practice-based ideology […] based on the values of sharing, common (intellectual) ownership and new social co-operations.”

Distinctive, yet with perhaps an interesting hint from “communism”, the term “Commonism” was first coined by Tom DeWeese, the president of the American Policy Center yet more recently redefined in a new book Commonism: A New Aesthetics of the Real edited by Nico Dockx and Pascal Gielen.

According to their introduction:

“After half a century of neoliberalism, a new radical, practice-based ideology is making its way from the margins: commonism, with an o in the middle. It is based on the values of sharing, common (intellectual) ownership and new social co-operations. Commoners assert that social relationships can replace money (contract) relationships. They advocate solidarity and they trust in peer-to-peer relationships to develop new ways of production.

“Commonism maps those new ideological thoughts. How do they work and, especially, what is their aesthetics? How do they shape the reality of our living together? Is there another, more just future imaginable through the commons? What strategies and what aesthetics do commoners adopt? This book explores this new political belief system, alternating between theoretical analysis, wild artistic speculation, inspiring art examples, almost empirical observations and critical reflection.”

In an interview excerpted from the book, author Pascal Gielen, Vrije Universiteit Brussel professor Sonja Lavaert, and philosopher Antonio Negri discuss how commonism has the ability to transcend the ideological spectrum. The commons, regardless of political leanings, collaborate to “[re-appropriate] that of which they were robbed by capital.” Examples put forward in the interview include “liberal politicians write books about the importance of the basic income; neonationalism presents itself as a longing for social cohesion; religiously inspired political parties emphasize communion and the community, et cetera.”

In another piece, Louis Volont and Walter van Andel, both of the Culture Commons Quest Office, argue that an application of commonism can be found in blockchain. They argue that Blockchain’s attributes are capable of addressing the three elements of the tragedy of the commons, which are “overuse, (absence of) communication, and scale”. Further, its decentralization feature enables a “common” creation of value.

Although, the authors caution of a potential tragedy of blockchain by asserting that:

“But what would happen when that one thing that makes the world go around – money (be it virtual, be it actual) – enters the picture? One does not need to look far: many cryptocurrencies, Bitcoin among them, are facilitated by blockchain technology. Even though it is ‘horizontally organized’, ‘decentralized’ or ‘functioning beyond the market and the state’, the blockchain-facilitated experiment of virtual money relates to nothing more than exchange value. Indeed, the core question one should ask when speculating on the potentialities of the blockchain experiment, is whether it is put to use for exchange value on the one hand, or for use value on the other. The latter, still, is where the commons begin. The former (that is, the imperatives of capital and its incessant drive for accumulation through trade), is where the blockchain mutates from a solution to a tragedy, to a comedy in itself.”

Mechanistic Evidence


/ˌmɛkəˈnɪstɪk ˈɛvədəns/

Evidence about either the existence or the nature of a causal mechanism connecting the two; in other words, about the entities and activities mediating the XY relationship (Marchionni and Samuli Reijula, 2018).

There has been mounting pressure on policymakers to adopt and expand the concept of evidence-based policy making (EBP).

In 2017, the U.S. Commission on Evidence-Based Policymaking issued a report calling for a future in which “rigorous evidence is created efficiently, as a routine part of government operations, and used to construct effective public policy.” The report asserts that modern technology and statistical methods, “combined with transparency and a strong legal framework, create the opportunity to use data for evidence building in ways that were not possible in the past.”

Similarly, the European Commission’s 2015 report on Strengthening Evidence Based Policy Making through Scientific Advice states that policymaking “requires robust evidence, impact assessment and adequate monitoring and evaluation,” emphasizing the notion that “sound scientific evidence is a key element of the policy-making process, and therefore science advice should be embedded at all levels of the European policymaking process.” That same year, the Commission’s Data4Policy program launched a call for contributions to support its research:

“If policy-making is ‘whatever government chooses to do or not to do’ (Th. Dye), then how do governments actually decide? Evidence-based policy-making is not a new answer to this question, but it is constantly challenging both policy-makers and scientists to sharpen their thinking, their tools and their responsiveness.”

Yet, while the importance and value of EBP are well established, the question of how to establish evidence is often answered by referring to randomized controlled trials (RCTs), cohort studies, or case reports. According to Caterina Marchionni and Samuli Reijula these answers overlook the important concept of mechanistic evidence.

Their paper takes a deeper dive into the differences between statistical and mechanistic evidence:

“It has recently been argued that successful evidence-based policy should rely on two kinds of evidence: statistical and mechanistic. The former is held to be evidence that a policy brings about the desired outcome, and the latter concerns how it does so.”

The paper further argues that in order to make effective decisions, policymakers must take both statistical and mechanistic evidence into account:

“… whereas statistical studies provide evidence that the policy variable, X, makes a difference to the policy outcome, Y, mechanistic evidence gives information about either the existence or the nature of a causal mechanism connecting the two; in other words, about the entities and activities mediating the XY relationship. Both types of evidence, it is argued, are required to establish causal claims, to design and interpret statistical trials, and to extrapolate experimental findings.”

Ultimately Marchionni and Reijula take a closer look at why introducing research methods that beyond RCTs is crucial for evidence-based policymaking:

“The evidence-based policy (EBP) movement urges policymakers to select policies on the basis of the best available evidence that they work. EBP utilizes evidence-ranking schemes to evaluate the quality of evidence in support of a given policy, which typically prioritize meta-analyses and randomized controlled trials (henceforth RCTs) over other evidence-generating methods.”

They go on to explain that mechanistic evidence has been placed “at the bottom of the evidence hierarchies,” while RCTs have been considered the “gold standard.”

Evidence Hierarchy — American Journal of Clinical Nutrition

However, the paper argues, mechanistic evidence is in fact as important as statistical evidence:

“… evidence-based policy nearly always involves predictions about the effectiveness of an intervention in populations other than those in which it has been tested. Such extrapolative inferences, it is argued, cannot be based exclusively on the statistical evidence produced by methods higher up in the hierarchies.”

Sources and Further Readings: