30 December 2014

BodyCams and other visibility

Three items on body cams and a defence by Microsoft apologists ...

'Moral Panics and Body Cameras' by Howard M. Wasserman in (2015) Washington University Law Review (Forthcoming) states that
This Commentary uses the lens of "moral panics" to evaluate public support for equipping law enforcement with body cameras as a response and solution to events in Ferguson, Missouri in August 2014. Body cameras are a generally good policy idea. But the rhetoric surrounding them erroneously treats them as the single guaranteed solution to the problem of excessive force and police-citizen conflicts, particularly by ignoring the limitations of video evidence and the difficult questions of implementing the body camera program. In overstating the case, the rhetoric of body cameras becomes indistinguishable from rhetoric surrounding responses to past moral panics.
'Visibility is a Trap: Body Cameras and the Panopticon of Police Power' by Eric Anthamatten claims that
One of the responses to the recent grand jury non-indictment in the death of Michael Brown was a call to equip police officers with cameras, the idea being that somehow this “third eye” will allow us to “see” the truth in a more objective way. If only we had a camera, we would know better what happened between Officer Darren Wilson and Michael Brown on that street. The human eye is not reliable, so we need a machine eye that is by its very nature disinterested and objective. The discussion of justice become not about systems and institutions of power, but conversations about vision, whether or not it is legal to film the police, whether or not it is a violation of our rights to have the police film us. If we can just police the police, watch the watchers, perhaps the asymmetry of power would be balanced or negated, and justice will somehow obtain. ...
For Foucault, the Panopticon became a symbol for “disciplinary society,” one that “called for multiple separations, individualizing distributions, an organization in depth of surveillance and control, an intensification and a ramification of power.” Power did not operate (only) by repression and overt force, but through these more subtle (and now, not so subtle) fragmentations that tear apart and recreate subjectivity and personhood, shape this “collection of separated individualities,” atomize and vaporize, a power that makes those “elementary particles” more manageable and docile. Disciplined bodies become “the object of information, never a subject of communication.”
“The Panopticon is a privileged place for the experiments on men,” “a kind of laboratory of power.”
On December 1st, amidst the varying levels of response to the non-indictment of Officer Wilson, President Obama requested $236 million to invest in body cameras and police training in order to restore trust in policing (nevermind that “trust” involves not having to watch someone at every moment). Two days later, a Staten Island grand jury decided not to indict Officer Daniel Pantaleo in the death of Eric Garner, an event that was caught on camera. Many, both liberal and conservative, clearly “saw” an injustice and an abuse power. Others saw Garner resist which, in their minds, justified the response by Pantaleo. Immediately, the “solution” of increasing cameras became problematic, if not farcical — even the visual evidence was not enough to indict, belying and underlying systemic problem that shapes the way we “see.”
But it is not simply a question of interpretation and how one “judges” the events, something that inevitably occurs in and through the double-interpretation of perception via any medium (text, photograph, video). Such a solution is a fetishization of sight that evades the underlying problem, a problem that not only has to do with race and class, but also the very structures, technologies and deployments of power in modern society ...
While Foucault provides a compelling analysis of the relationship between surveillance, discipline, and the deployment of power, what he’s articulating is something that is experienced daily by people of color in the United States, namely the experience of constantly being watched while moving through public space, of being always marked by skin color, manner of dress, or physical comportment, what W.E.B. Dubois calls a “double-consciousness.” It is the experience of not only being a “suspicious” body, but of being disciplined, controlled, and already indicted in and through those surveilling eyes. It is the expiring of being incarcerated, unfree.
“The Panopticon ... must be understood as a generalizable model of functioning; a way of defining power relations in terms of the everyday life of men.” Yes, we all live in a Panopticon. But it is not only the Panopticon of Bentham or Orwell, a central tower from which the gaze operates. Rather, it is the Panopticon of Kafka, one that is everywhere precisely because there is no centralization, where we, the surveilled, are also the surveillers, we the watched are also the watchers. “Consequently, it does not matter who exercises power. Any individual, taken almost at random, can operate the machine: in the absence of the director, his family, his friends, his visitors, even his servants.”
Such surveillance has become normalized and distributed, into our own pockets, onto our own bodies. It is not a great leap to imagine the police outfitted with, alongside their peppery spray and pistols, glasses that record everything, or perhaps even cameras embedded into their own eyes. Is this the image of justice and freedom? Will this protect the citizenry and help to reduce racism, classism, and abuses of power?
Perhaps surveillance will help both police officers and citizens feel more secure because they feel they will be accountable to some disinterested third party or to the “court” of public judgment. There is some recent evidence that use of force declines when body cameras are present. But, as Foucault emphasizes, surveillance is yet another refinement of power and control, a technology, however well-intentioned, that continues to atomize our bodies in time and space as a way of examining, fragmenting, and controlling those bodies. There is no justice “behind” the way we realize it through our technologies and systems. These cameras, then, do not become the tool of justice, but a catalyst for surveillance, discipline, and punishment. The camera replaces the gun — the violence and control over a body is no less totalizing.
“Broken windows” leads to broken windows. The “riot” is, at some level, an expression of exclusion from property and meaningful participation and recognition in the life of society. Many see it as a breaking in, but it is in fact a breaking out of the “dungeon” of surveillance and control perpetuated by modern biopower. This is something that bodies that are not under siege do not and perhaps cannot understand. From the safety of their own “Panopticon,” behind the glass of the television, in the comfort of their living room chair, they watch these “animals” and only see “thugs,” “hoodlums,” “criminals,” a “prison riot,” not to mention other choice labels by which they “see” these bodies.
This is precisely the point: poor communities where most of the bodies are brown experience “public” and “free” space as surveilled space, controlled space, a space where their bodies are not their own but perpetually disciplined, fragmented, and examined by the various modes of power. Are more eyes the answer?
Visibility is a trap
'Are You Recording This?: Enforcement of Police Videotaping' by  Martina Kitzmueller in (2014) 47(1) Connecticut Law Review comments
Increasing numbers of police departments equip officers with dashboard or body cameras. Advances in technology have made it easy for police to create and preserve videos of their citizen encounters. Videos can be important pieces of evidence; they may also serve to document police misconduct or protect officers from false allegations. Yet too often, videos are lost, destroyed, or never made, often depriving criminal defendants of the only objective evidence in a case. When this happens, there is not always a consequence to the prosecution. This Essay explores this problem of enforcement by examining how different states are compelling law enforcement to make and preserve videos through a combination of legislation and judicial intervention.
'Do-Not-Track and the Economics of Third-Party Advertising' (Boston University School of Management Research Paper No. 2505643) by Ceren Budak, Sharad Goel, Justin M. Rao and Georgios Zervas argues that
 Retailers regularly target users with online ads based on their web browsing activity, benefiting both the retailers, who can better reach potential customers, and content providers, who can increase ad revenue by displaying more effective ads. The effectiveness of such ads relies on third-party brokers that maintain detailed user information, prompting legislation such as do-not-track that would limit or ban the practice. We gauge the economic costs of such privacy policies by analyzing the anonymized web browsing histories of 14 million individuals. We find that only 3% of retail sessions are currently initiated by ads capable of incorporating third-party information, a number that holds across market segments, for online-only retailers, and under permissive click-attribution assumptions. Third-party capable advertising is shown by 12% of content providers, accounting for 32% of their page views; this reliance is concentrated in online publishing (e.g., news outlets) where the rate is 91%. We estimate that most of the top 10,000 content providers could generate comparable revenue by switching to a “freemium” model, in which loyal site visitors are charged $2 (or less) per month. We conclude that do-not-track legislation would impact, but not fundamentally fracture, the Internet economy.

28 December 2014

Social sorting

The lucid 90 page The Scoring of America: How secret consumer scores threaten your privacy and your future  [PDF] by Pam Dixon and Robert Gellman for the World Privacy Forum comments
To score is human. Ranking individuals by grades and other performance numbers is as old as human society. Consumer scores — numbers given to individuals to describe or predict their characteristics, habits, or predilections — are a modern day numeric shorthand that ranks, separates, sifts, and otherwise categorizes individuals and also predicts their potential future actions.
Consumer scores abound today. Credit scores based on credit files receive much public attention, but many more types of consumer scores exist. They are used widely to predict behaviors like, spending, health, fraud, profitability, and much more. These scores rely on petabytes of information coming from newly available data streams. The information can be derived from many data sources and can contain financial, demographic, ethnic, racial, health, social, and other data.
The Consumer Profitability Score, Individual Health Risk Score, Summarized Credit Statistics that score a neighborhood for financial risk, fraud scores, and many others seek to predict how consumers will behave based on their past behavior and characteristics. Predictive scores bring varying benefits and drawbacks. Scores can be correct, or they can be wrong or misleading. Consumer scores – created by either the government or the private sector – threaten privacy, fairness, and due process because scores, particularly opaque scores with unknown ingredients or factors, can too easily evade the rules established to protect consumers.
The most salient feature of modern consumer scores is the scores are typically secret in some way. The existence of the score itself, its uses, the underlying factors, data sources, or even the score range may be hidden. Consumer scores with secret factors, secret sources, and secret algorithms can be obnoxious, unaccountable, untrustworthy, and unauditable. Secret scores can be wrong, but no one may be able to find out that they are wrong or what the truth is. Secret scores can hide discrimination, unfairness, and bias. Trade secrets have a place, but secrecy that hides racism, denies due process, undermines privacy rights, or prevents justice does not belong anywhere.
Broader transparency for consumer scores with limited secrecy may offer a middle ground. Knowing the elements but not necessarily the weights of a scoring system provides a partial degree of openness and reassurance. Knowing that there is a scoring system and how and when it is used helps. Knowing the source and reliability of the information used to make a score helps. Being able to challenge a score and correct the data on which it is based helps. Knowing that some types of information will not be used for scoring helps. Knowing that data collected for one purpose will not be used for another or in violation of law helps. Knowing that the person running the scoring system is accountable in a meaningful way helps.
The history of the credit score provides a useful model for the new batch of predictive consumer scores. Developed in the 1950s, the credit score became part of consumer credit granting. The credit score was largely secret to the consumers that it scored and affected until 2000, when a long and well-documented history of unfair uses and abuses finally culminated in the credit score being made available to consumers. Eventually, public pressure caused the credit score’s use and even its underlying factors to become public. The use of factors such as race, gender, and religion were prohibited and this was spelled out in detail in law.
No similar protections exist for most consumer scores today. Consumer scores are today where credit scores were in the 1950s. Data brokers, merchants, government entities, and others can create or use a consumer score without notice to consumers. For various reasons laws governing credit scores do not typically extend protection to the new consumer scores. We need rules that will make consumer scores fair, accountable, accurate, transparent, and non-discriminatory.
This report discusses and explores consumer scores, what goes into them and how they are made, how they are used, the regulations in place that control some but not most new consumer scores, and how scores affect broader privacy and fairness issues. The discussion of findings and recommendations points toward solutions and reforms that are needed.
The report is structured as follows
Part I: Summary and Background
As the numbers of predictive consumer scores increase and their usage expands, Americans face a future that may be shaped in significant ways by consumer scores. By itself, consumer scoring is not necessarily good or bad. Scoring orders consumers along a mathematically defined scale. However, scoring has the prospect of being used to affect individuals in significant ways that may not always be fair or even legal. If a predictive score unknown to a consumer determines how that consumer is treated, the results may not be acceptable to the American public. The quality and relevance of the data used, the transparency of the methodology of how the score was created, plus the reasonableness of the application of the consumer score are the major factors that determine the fairness of any scoring activity. These issues should be the central focus of any policy debate about consumer scoring. These issues also suggest the elements of best practices that should apply to consumer scoring.
What is a Consumer Score?
With this report, the World Privacy Forum introduces a term: consumer scores. Consumer scores – the ones we discuss in this report – are built using predictive modeling. Predictive modeling uses copious amounts of information fed through analytical methods to predict the future, based on past information. Predictive consumer scores are important because they affect the lives, privacy, and wellbeing of individuals. Many people know about credit scores, but few know about the broader range of new consumer scores. Consumer scores are already abundant and are in active use. Consumer scores are not just an online phenomenon. Consumer scores are found in a wide array of “offline” arenas, including businesses, health care providers, financial institutions, law enforcement, retail stores, federal and state government, and many other locations. Some social consumer scores may have online applications, but mostly, consumer scores are not solely focused on just online activities. And unlike credit scores, consumer scores remain largely secret and unregulated.
The World Privacy Forum defines a consumer score as follows: A consumer score that describes an individual or sometimes a group of individuals (like a household), and predicts a consumer’s behavior, habit, or predilection. Consumer scores use information about consumer characteristics, past behaviors, and other attributes in statistical models that produce a numeric score, a range of scores, or a yes/no. Consumer scores rate, rank, or segment consumers. Businesses and governments use scores to make decisions about individual consumers and groups of consumers. The consequences can range from innocuous to important. Businesses and others use consumer scores for everything from predicting fraud to predicting the health care costs of an individual to eligibility decisions to almost anything.
Who has a Score?
Consumer scoring is already more widespread than most people realize. Many hundreds of consumer scores exist, perhaps thousands. How many Americans have them? Almost all do. Minors are less likely to be scored than adults, although they, too can have or influence some consumer scores. For example, household scores often reflect interests and activities of minors.
Among American adults, each individual with a credit or debit card or a bank account is likely to be the subject of one or more scores. Many individuals signed up under the Affordable Care Act have a score. Individuals who buy airline tickets have a score. Individuals who make non-cash purchases at large retail stores likely have a score. Scores such as the medication adherence score, the health risk score, the consumer profitability score, the job security score, collection and recovery scores, frailty scores, energy people meter scores, modeled credit scores, youth delinquency score, fraud scores, casino gaming propensity score, and brand name medicine propensity scores are among the consumer scores that score, rank, describe, and predict the actions of consumers.
In short, almost every American over the age of 18 has at least one score, and most adult Americans have many scores. An individual could easily be the subject of dozens or even hundreds of secret consumer scores. We can safely predict that there will be many more consumer scores in the future. Fed by the masses of consumer data now available, consumer scoring is quickly becoming a form of shorthand to make sense of a sea of information.
Gaps in Consumer Privacy Rights and Protections around Consumer Scoring (And why existing laws don’t always apply)
This report’s analysis is that many new consumer scores exist, and many of these new scores do not appear to fall under the narrow protections offered by the Fair Credit Reporting Act or the Equal Credit Opportunity Act for a variety of reasons. Scores built from factors outside a formal credit bureau file, scores designed to predict the behavior of groups of people instead of individuals, and new scores in emerging and unregulated areas may all fall outside of existing protections.
For example:
• Energy consumption scores, churn scores, and identity scores are not likely to fall under the FCRA and other laws as currently written. This is because those scores do not meet the layers of qualifications that would bring them under the FCRA.
• Scores that identify the approximate credit capacity of neighborhoods instead of individuals also appear to be unregulated. This is because the FCRA applies to individuals, not neighborhoods. Formal credit scores may only be used in certain circumstances, for example, for extending a firm offer of credit or insurance. Credit scores cannot be used for general marketing purposes, but aggregate credit statistics tied to a neighborhood do not appear to be subject to the same restrictions for the reasons mentioned. Lead generation is not the same thing as a formal offer of credit under the FCRA.
• Risk scores -- like health risk scores -- that use broad demographic information and aggregate financial statistics about consumers to assess financial or other risks (credit bureau files are not typically used) also don’t appear to fall under the layers of requirements that would bring them under current regulation.
• The Equal Credit Opportunity Act requires credit scoring systems to not use race, sex, marital status, religion, or national origin as factors comprising the credit score. But this law applies only to what is today a narrowly defined credit scoring system. Other scores which fall outside of the narrow definitions – like identity, fraud, churn, and other predictive scores can incorporate factors that would in other situations be considered prohibited factors to use.
As a result, consumers may have scant rights to find out what their non-FCRA consumer scores are, how the scores apply to them and with what impact, what information goes into a score, or how fair, valid, or accurate the score is. Even if the input to a score is accurate, consumers do not know or have any way to know what information derived from their lifestyle, health status, and/or demographic patterns is used to infer patterns of behavior and make decisions that affect their lives.
Further, consumers can have difficulty exercising basic Fair Information Principles for many if not most new consumer scores. Fair Information Principles form the base for most global privacy law today, including some US privacy laws. However, those who create unregulated scores have no legal obligation to provide Fair Information Practices or due process to consumers.
These significant gaps in consumer protections mean that consumer scores may include or use discriminatory factors in their composition, or uncorrected or otherwise inaccurate information could be included. Scores developed to characterize individuals or predict their behaviors need to provide fairness and due process. The credit score is already subject to some regulation, but that is not to say that consumers would not benefit from better rules for credit scores.
There is a great need to examine the effects and fairness for all consumer scores now in use. Intriguing possibilities exist that a certain stratification of consumer experience based on opportunities offered to each consumer could become commonplace. Victims of identity theft, for example, may consistently receive different and less desirable marketing treatment than individuals with clean credit scores, even if most other demographic factors are similar.
Disparate treatment, even in the area of marketing opportunities granted to consumers, raises many questions, questions that the general field of risk-based pricing has raised. Oddly, direct marketing lists and activities have the potential to strike deeply into the lives of individuals in quirky ways that can have an impact on consumer lifestyle. Much remains to be learned about the impact of consumer scoring in the direct marketing arena, as well as eligibility issues and edge-eligibility issues like scores for identity and authentication.
It identifies a range of issuere regarding gaps in protection and information fairness in scoring -
Key Issue: Score Secrecy
There are good reasons why credit scores are not secret anymore, nor are the foundational factors that comprise the score. By law, consumers have the right to see credit scores now. This report finds that with the exception of the credit score and a handful of other consumer scores, at this time, secrecy is the hallmark of many consumer scores. The factors that go into most scores are usually secret, the models used are usually secret, and in many cases, the score itself is also secret. This report’s analysis is that consumer scores that are risk scores bear many similarities to scores regulated under the FCRA. Yet industry treats these risk scores as falling outside the FCRA so that consumers have none of the rights guaranteed by the FCRA.
Consumers have no formal rights to find out what their non-FCRA consumer scores are, or how these scores affect their lives. Victims of identity theft and other individuals may have errors or omissions affecting their scores, but they do not necessarily have a right to see or correct the scores. Even if information is accurate, consumers do not know or have any way to know how companies use information derived from their lifestyle, health status, and/or demographic patterns to infer patterns of behavior and make decisions that affect their lives. Unseen scores can affect consumers’ marketplace experiences and much more.
Key Issue: Score Accuracy
Because consumers do not have the right to correct or control what personal information goes into a consumer score as an attribute or factor, the accuracy of the scores is suspect. Consumers also do not have the right to see the scoring models used to make the score, nor do they typically have information about the model validity. Because of the lack of transparency, consumers cannot be assured of the reliability, fairness, or legality of scoring models. Inaccurate, incomplete, and illegal factors may be used today to make decisions about consumers without any oversight or redress.
Credit scores offer a useful model here. Credit scores are based on credit reports. Credit report accuracy has been the subject of substantial, meaningful scrutiny over decades. The CFPB, in its 2012 study on credit reports, noted the significant problems with inaccuracy that occur:
Given the volume of data handled, the challenges of matching tradelines to the correct consumer files, and the number and variety of furnishers, inaccuracies in some credit files inevitably occur. Inaccuracies in credit files and credit reports can occur where information that does not belong to a consumer is attached to his or her file, where information belonging to a consumer is omitted from the file, or where there are factual inaccuracies in trade line or other information in the consumer’s file. Some of these inaccuracies can be attributed to matching challenges in assigning a trade line to a consumer’s file. Other causes of inaccuracies include data and data entry errors, NCRA system or process inaccuracies, furnisher system or process inaccuracies, identity fraud, or time lags.
In a ten-year, Congressionally-mandated study published in 2013, the FTC found that overall “one in five consumers had an error on at least one of their three credit reports.” The FTC found that these credit report errors did impact the credit score. The FTC found that, specifically:
• “Slightly more than one in 10 consumers saw a change in their credit score after the CRAs modified errors on their credit report; and;
• Approximately one in 20 consumers had a maximum score change of more than 25 points and only one in 250 consumers had a maximum score change of more than 100 points.”
Errors in credit scores abound, and credit scores are based on credit reports, which also are subject to significant errors. If a transparent score with few factors has these kinds of errors, what about consumer scores? Consumer scoring relies on dozens, hundreds, or thousands of data elements that have no standards for accuracy, timeliness, or completeness. The quality of data matters: errors in data used to make a score create a score that is not predictive. With thousands of factors, error rates and false readings become a big issue.
Key Issue: Identity Theft and Consumer Scoring
Victims of identity theft – both financial and medical forms of the crime -- may have significant and stubbornly ongoing errors or omissions affecting their scores. ID theft victims can be seriously affected by identity scoring because their identity scores and fraud scores may be incorrect as a direct result of criminal activity. This can cause a range of problems from being denied services to being tagged as a potential fraudster. Yet even this vulnerable group has no right to see or correct many consumer scores.
Key Issue: Unfairness and Discrimination
One of the fundamental policy issues regarding scoring activities is the question of what characteristics it is appropriate to use in scoring consumers. In the world of home loans, ECOA has answered that question. But in the world of direct marketing, this area is nearly without boundaries. In a prescient early critique of scoring policy, Columbia University professor Noel Capon wrote in 1982:
Since prediction is the sole criterion for acceptability, any individual characteristic that can be scored, other than obviously illegal characteristics, has potential for inclusion in a credit scoring system.
As a bewildering plethora of new databases of consumer information become available, these databases may be scored in various ways by being run through one or more scoring models. More databases of consumer information fundamentally can mean more potential scores, and more potential characteristics to score.
The Equal Credit Opportunity Act protects consumers from invidious discrimination in formal credit granting situations. Notably, the ECOA requires that credit scoring systems may not use race, sex, marital status, religion, or national origin as factors comprising the score. The law allows creditors to use age, but it requires that seniors be treated equally. But in the modern consumer scores, marital status – a protected factor under ECOA – is commonly used as a consumer score factor. Consumer scores may also contain underlying factors of race, sex, and religion without disclosure to consumers. In some cases, health factors may also be included in scores, for example, if a person smokes, or has a chronic illness. (See the section on Factors below for an example of a score that incorporates smoking and ethnicity).
As discussed in Part II of this report, a single score is often created from the admixture of more than 600 to 1,000 to even 8,000 individual factors or data streams. These factors can include race, religion, age, gender, household income, zip code, presence of medical conditions, zip code + 4, transactional purchase information from retailers, and hundreds more data points about individual consumers. Therefore, one individual score can have the potential to contain hidden factors that range from bland – like mail order buyer of sports goods -- to quite sensitive – like ethnicity.
A score designed to assess or assign consumer value to a business could easily include factors that would be entirely unacceptable or that, in the context of either the Equal Credit Opportunity Act (ECOA) or the Fair Credit Reporting Act, would be flatly illegal. If ECOA factors are present in consumer scores, in most cases it would be difficult or impossible for consumers to find out if the scoring system or its factors were secret. While carefully directed and controlled use of credit scoring and credit automation has reduced some discriminatory practices, new consumer scoring that uses elements that correlate with prohibited factors such as race can reintroduce discrimination and hide the effects behind a secret or proprietary screen that falls entirely outside of current consumer protection regulations. This is not acceptable.
Key Issue: Sensitive Health and Lifestyle Information and Consumer Scoring
Health scores already exist. This category of score deserves special attention and scrutiny. Some health scores are used in the HIPAA context, some are used outside the HIPAA context. The health scores used outside the HIPAA context are of most concern. Actuaries already use some new consumer scores to underwrite risk, for example, the Brand Name Medicine Propensity Score from a health category and the Underbanked Indicator from the financial category. Scores can contain health information as hidden information within the score, and used for health purposes, or used for non-health related purposes such as marketing, or risk scoring. Many consumers with chronic health conditions would object strenuously to having their financial risk be determined by their health status. While health risk may be very predictive in a score, is it fair to use without consumer knowledge?
Just because a score contains information about a consumer’s health status, it does not mean the score will be subject to the federal health privacy rule (HIPAA). In fact, much of health information available for commercial use outside of the healthcare environment falls outside the scope of HIPAA. HIPAA, for example, provides no consumer privacy rights over health data held by list and data brokers. Health information often leaks outside of HIPAA protections when it is revealed by consumers through surveys, website registrations, and other online activities. After a consumer reveals his or her health information to a non-HIPAA third party, that information is considered out of HIPAA’s bounds. It is in this way that consumers’ most sensitive health information can wind up used as fodder for a consumer score, with unknown consequences.
Consumer scores that use health or other sensitive information such as sexual orientation as factors need close examination for fairness, and consumers need rights over whether their health information is used in predictive scores, whether for marketing or any other purpose.
Key Issue: Consent and Use of Consumer Data in Predictive Scores
If a consumer fills out a registration on a health-related web site or a consumer warranty card that accompanied a purchase, the consumer did not give informed consent that the information can be used in downstream consumer scoring in ways that affect the consumer’s marketplace opportunities. A buried statement in an unread privacy policy that “we may share your information for marketing purposes with third parties” is not informed consent to allow unfettered use information for predictive scoring. Does making a purchase with a credit or debit card at a retailer grant consent for use of a consumer’s purchases and other information to be used in a score? Part II of this report contains a detailed discussion of what kinds of information go into consumer scores. Many individuals would be quite surprised to learn just how the details of their lives are fodder for scores they may never see or have access to – and did not knowingly consent to. The issue of consent becomes increasingly important for scores that affect any kind of eligibility, such as jobs, credit, insurance, identity verification, or other significant opportunities.
Scores Then: A Handful of Factors. ……Scores Now: Thousands of Factors
The research for this report found that consumer scores may rely on hundreds or thousands of pieces of consumer information coming from many different data sources. This report identifies a large roster of raw consumer data that includes demographic information like age, race, gender, ethnicity, and home address as well as religion, mobile phone number, online and offline purchase history, health conditions like Alzheimer’s, diabetes, and multiple sclerosis, as well as intimate financial details such as net worth, card holder information, low or high-end credit scores, money market funds, ages of children, and a great deal more.
Statistical scoring methods rely on the increasing availability of large amounts of new source data from social media, the web in general, and elsewhere. The input for consumer scores can include information that is mostly unobjectionable or public. But, as discussed, consumer scores also can incorporate highly sensitive information that in other contexts could be used in a prejudicial, unfair, or unethical way in making decisions about consumers. Some data, such as social media data, can be unobjectionable in one context, but inappropriate as a factor, for example, in credit decisioning models. An example, and a fairly common one, is of a predictive model that a major US health insurer worked with an analytics company to create. The idea was to determine whether or not publicly available consumer data could enhance the quality and effectiveness of their predictive risk models. They tested approximately 1,500 factors at the household level and found that the consumer information that showed the most value in predicting individual level risk included:
• Age of the Individual
• Gender
• Frequency of purchase of general apparel
• Total amount from inpatient claims
• Consumer prominence indicator
• Primetime television usage
• Smoking
• Propensity to buy general merchandise
• Ethnicity
• Geography – district and region
• Mail order buyer - female apparel
• Mail order buyer - sports goods - Those unfamiliar with predictive models can find it surprising to learn that information about purchasing sporting goods can become a part of a predictive score for a health insurer. But the factors used in this example are not surprising factors to find in a modern predictive consumer score model. This is actually a fairly short list compared to some models with thousands of factors.
The raw source material for the factors fed into consumer scores comes from sources such as:
• Retailers and merchants via Cooperative Databases and Transactional data sales & customer lists.
• Financial sector non-credit information (PayDay loan, etc.)
• Commercial data brokers
• MultiChannel direct response
• Survey data, especially online
• Catalog/phone order/Online order
• Warranty card registrations
• Internet sweepstakes
• Kiosks
• Social media interactions
• Loyalty card data (retailers)
• Public record information
• Web site interactions, including specialty or knowledge-based web sites
• Lifestyle information: Fitness, health, wellness centers, etc.
• Non-profit organizations’ member or donor lists
• Subscriptions (online or offline content) (Part II of the report contains a more complete list.)
Traditionally, much of this data came from data brokers or mailing list sellers. That is still the case, but now many new data streams are now available. So-called big data (large data sets) is one source. Other new data, particularly mobile and social data streams, comes via application programming interfaces (API). Data sets that used to be too large for all but the largest of companies to handle computationally can now be replicated and massaged by smaller firms and dedicated analytics teams within companies. Small analytics companies now compete with large data brokers to offer predictive analytics as well as data. One company states they use 300 billion data attributes in compiling their predictive scores, compiled from 8,000 data files. This is no longer an extraordinary feat, it is competitive and to be expected in a world with large data flows.
Analytics tools will continue to come down in price, just as consumer data has become a commodity item. Widespread and inexpensive data and analytics have the potential to allow broader use of predictive analytics. Consumer scores may proliferate, especially in the absence of any need for accuracy, fairness, or transparency. Consumer scoring may expand just because it is cheap and fashionable. Merchants themselves may have little ability to judge the accuracy of consumer scoring.
In short, given abundant data and more data tools, factors used to create consumer scores could continue to increase. With each new unverified factor comes the risk of extra errors or unfairness due to sensitive or prejudicial or irrelevant factors.
Scoring Methods and Models are Opaque
To create a consumer score, the score modeler feeds raw information (factors about consumers) into an algorithm designed to trawl through reams of data to detect consumer behavior patterns and to eventually sift consumers into a ranking by their scores. Each score generally has a name and predictive or descriptive function. Credit scores are the best-known example of this. With credit scores, information culled from a consumer’s credit bureau file becomes the raw input into a formal credit scoring model. Credit scores are built on credit file data. There is a nexus between the score and the data. The data is intrinsic to the score. The Fair Credit Reporting Act lays out, in concert with the Equal Credit Opportunity Act, a variety of responsibilities and restrictions in the uses of credit report data to use in credit scores. It is a balanced approach, and the Fair Credit Reporting Act remains a strong privacy law that enables Fair Information Practices for consumers.
Today, though, scoring models are easily built from data that is extrinsic to the final score. No nexus may exist between the input to a score and the output. In the financial scoring area, companies can now build financial scores from social media, demographic, geographic, retail purchase history, and other non-traditional information that may not be included in the formal credit file. In the health arena, analysts can now build health risk scores from mere wisps of demographic data, without any actual patient records. In this is a new world of scoring, where analysts use factors extrinsic to the purpose of the score to build scores, that a person has red hair can be used as a factor. And the more factors, the better. Instead of using 30 factors, why not 3,000?
The use of credit information for pricing insurance risk is an example of this. Statisticians and actuaries predict the cost of providing car or homeowners insurance using selections of credit report factors about a driver or homeowner. These insurance scores reportedly have a predictive capability. Yet there is no overt reason why credit worthiness correlates with the risk involved with driving safety. There is much controversy about the use of a statistical correlation that does not appear to be causal. Some states restrict the use of credit information for insurance, but the practice remains common.
Many of the consumer scores discussed in Part III are new classes of scores. When scores have hundreds and thousands of factors, it stands to reason that a causal link becomes much more tenuous. The more factors, the less casual the link may be. Risks associated with models are discussed in Part II of the report.
Examples and Numbers of Consumer Scores
Consumer scoring is growing. In 2007, the research for this report uncovered less than 25 scores. In 2014, the research uncovered hundreds of scores, with the strong likelihood that thousands of custom scores exist beyond our ability to confirm them. Here are some examples of consumer scores: Consumer profitability scores predict, identify, and target marketing prospects in households likely to be profitable and pay debt. The Job Security Score claims to predict future income and capacity to pay. Churn scores seek to predict when a customer will move his or her business or account to another merchant (e.g., bank, cell phone, cable TV, etc.) The Affordable Care Act (ACA) health risk score creates a relative measure of predicted health care costs for a particular enrollee. In effect, it is a proxy score for how sick a person is. The Medication Adherence Score predicts if you are likely to take your medication according to your doctor’s orders. Brand Name Medicine Propensity Score – will you be purchasing generics or brand name medications? Fraud Scores indicate that a consumer may be masquerading as another, or that some other mischief is afoot. These scores are used everywhere from the Post Office at point of sale to retailers at point of sale to behind-the-scenes credit card transactions. This is a very widely used score, and a number of companies compete in the fraud score area. Part III of the report discusses these and other specific scores in detail.
Uses of Consumer Scores, Regulation, and Modern Eligibility
After a consumer is scored, ranked, described, or classified, companies, governments, private enterprises, health care entities, and others including law enforcement, can then use the resulting score to make decisions about an individual or group. This is why scores impact consumers every day.
Scores are gaining footholds as part of routine business processes for an expanding number of purposes for everything from marketing to assessing a person’s identity to predicting a person’s likelihood to commit fraud, and more. The consumer score acts as a form of predictive evaluation to measure, predict, and generally facilitate making a decision about things such as an individual’s:
• Credit worthiness, • Popularity, • Reputation, • Wealth, • Propensity to purchase something or default on a loan, • Measure health, • Measure/predict likelihood to commit fraud, • Measure/predict identity • Measure/predict energy consumption • Job success probability • Etc.
In the traditional credit score realm, some have argued that scoring is a foundational activity in the credit market, as well as a wholly positive factor. Others have said that the sub-prime meltdown of the late 2000’s was fueled by overreliance on scoring products. Some of the reasons for using credit scores have the potential to be helpful directly to consumers. Better and faster credit decisions help consumers, for example.
In new consumer scoring, some have argued that the scores are mainly just for marketing and are largely beneficial. There can be potential benefits for consumers, For example, consumer risk scores that prevent fraud are helpful up to a point. But any potential benefits are real only if the scoring models are correct and non-discriminatory, the data is timely, and the scores are something that consumers want. Credit score regulation provides transparency and imposes some limits on use and construction. That offers some assurance to consumers. But when other consumer scores enter the marketplace without transparency or the limits that apply to credit scoring, consumer benefits are much more uncertain, and unfairness is more likely.
There is continuum of concern regarding consumer scores. Some scores are used for straightforward marketing purposes. These scores may be of less concern (however the fairness of factors and secrecy and validity are still a concern). Of greater concern are the consumer scores that are used for what we call “modern eligibility.” This includes identity verification and fraud assessment scores, as well as credit decisioning scores and scores that are used to predict job success or decide between job applicants. These scores are especially worrisome because errors in these scores could lead to significant deleterious consumer impacts.
Whether a consumer receives a coupon for a free soda is not a big deal. In comparison, whether a consumer can complete a transaction is of significant consequence. Any score used for eligibility – like being approved for credit or a job -- becomes important. The most casual social scores meant just to measure social reach have on occasion been used as a criterion for judging applicant hiring qualifications, so all scores need to be explored and assessed.
Some scores –for example, aggregate credit scores not subject to the FCRA – can determine a neighborhood’s general credit score or range. Opportunities for individuals living in that neighborhood will be affected in ways that they cannot anticipate and in ways that bear no relationship to their personal situation. Forms of redlining – the practice of turning someone down for a loan or insurance because they live in an area deemed to be high risk – is a threat in these situations.
By all appearances, consumer scoring has sped beyond the old constraints that were imagined in a largely analog era, and real consumer harms can be the result.
Deja Vu: Why the History of the Credit Score is Important
History is repeating itself with consumer scoring. Before secret predictive consumer score issues, there were secret credit score issues. Credit scores had many of the same problems: secrecy, unfairness, inaccuracy, and opacity. Part IV of the report contains a detailed history of the credit score, including how that score became public and how consumers got important rights regarding credit scores and reports. In brief here, credit scores were unknown to most consumers through the 50s, 60s, 70s, and 80s. Trickles of a score that was not disclosed to consumers but that could be used to deny a person credit began to leak out slowly to some policymakers, particularly around the time ECOA passed. In May 1990, the Federal Trade Commission failed to protect consumers when it wrote commentary indicating that risk scores (credit scores) did not have to be made available to consumers. But when scoring began to be used for mortgage lending in the mid 90s, many consumers finally began hearing about a “credit score,”  many of them for the first time, and mostly when they were being turned down for a loan. A slow roar over the secrecy and opacity of the credit score began to build. By the late 90s, the secrecy of credit scores, the underlying methodology or factors that went into the score, and the scoring range became a full-blown policy issue. Beginning in 2000, a rapid-fire series of events – particularly the passage of legislation in California that required disclosure of credit scores to consumers – eventually ended credit score secrecy. Now, credit scores must be disclosed to consumers, and the context, range, and key factors are now known. This is an example of how brave State privacy legislation serves as a model for state and federal policy makers. In this case, the US “laboratory of democracy” took state legislation and turned it into a federal rule that protects consumers everywhere.
Credit scores are no longer secret anywhere in the United States, and this was and still is the right policy decision. Why are other scores used for important decisions about consumers still secret? Why do score factors and numeric ranges remain secret, when the risk of the data comprising the score of a factor used in modern eligibility practices such as identity verification or fraud identification is very high?
Consumer scores stand today where credit scores stood in the 1950s: in the shadows. While there are some happy exceptions to this, such as most social scores and a few other consumer scores, most consumer scores are not available for consumers to see. As a result, consumers have little to no ability to learn when their lives are affected in a major or minor way by a consumer score that they never heard about. Credit scores are not perfect and still present some issues, but we have learned much from the credit score. What we have learned most of all is that there should be no secret consumer scores and no secret scoring factors. If a score is being used in any meaningful way in a consumer’s life, he or she needs to know about it and have some choices regarding that score.
 Historically, some known consumer issues with the credit score include the following:
• Credit scores reflect inaccuracies in the credit reports they are based on, and credit reports have repeatedly been found to contain errors. • Victims of ID theft can experience changed credit scores. • Consumers who experience major life events such as medical events or divorce can pay a long price in the scoring world. • The FTC has brought cases around “mission creep” in the use of credit score outside of its regulated uses. (Credit scores may only be used for firm offers of credit or insurance, not for general marketing use.)
Findings by the authors are as follows -
Consumer scores are expanding in type, number, and use because of the growth of predictive analytics and the ready access to hundreds and thousands of factors as raw material. Just as credit scores were secret for decades until state and federal legislation mandated that consumers could see their credit scores, today consumer scores are largely secret. While new scores multiply, consumers remain in the dark about many of their consumer scores and about the information included in scores they typically don’t have the rights to see, correct, or opt out of. A primary concern is how these scores affect individuals and meaningful opportunities available to them. Another area of concern is the factors used in new consumer scores, which may include readily commercially available information about race, ethnicity, religion, gender, marital status, and consumer-reported health information.
This report’s other key findings are:
• Unregulated consumer scores – as well as regulated credit scores – are both abundant and increasing in use today.
• The information used in consumer scores comes from a large variety of sources. Some scores use thousands of factors or consumer attributes.
• Many consumer scores, the ranges of the scores, and the factors used in them are secret.
• A consumer score may, without any public notice, rely on an underlying factor or attribute that has discriminatory implications (e.g., race or gender) or that most consumers consider sensitive (e.g., health or financial).
• Consumer scores in use today affect a consumer’s marketplace opportunities. Some of these opportunities are major (e.g., financial, employment, health,), some are minor (e.g., receiving a coupon, spam, or junk mail), and many are in between. Consumers are adversely affected by scores that are kept secret, and consumers are adversely affected when they do not have rights to correct scores.
• Consumer scores are found in a wide array of “offline” arenas, including businesses, health care providers, financial institutions, law enforcement, retail stores, federal and state government, and many other locations. Some of the more social consumer scores may be online, but mostly consumer scores are not solely focused on just online activities.
• Consumers usually have no way to know what the scores predict or how the scores are used.
• Consumers typically have no notice or knowledge about the data sources used in scores predicting their behavior or characterizing them. Consumers typically have no rights over the data about themselves, and consumers usually have little to no ability to control use of the data.
• Consumers typically do not have the right to opt out of being the subject of a consumer score or to prevent use of a consumer score.
• Except where the Fair Credit Reporting Act applies to a consumer score, most consumer scores are not subject to any regulation for privacy, fairness, or due process. A lack of transparency makes it difficult or impossible to determine if creation or use of the scores violates a law that prohibits discrimination.
• Consumers who are victims of identity theft can have their credit or consumer scores affected thereby and may have little recourse even though errors may have major consequences for their ability to function in the economic marketplace can be major. Other consumers can also have their lives affected by the use of consumer scores to determine eligibility for important opportunities in the marketplace. Some consequences may be less significant.
• Consumers have remedies under state and federal law with respect to correcting and seeing their credit reports, but not necessarily with respect to the many records that contribute to consumer scores. Secret consumer scores do not provide consumers with correction rights of underlying information.
 The report offers several Key Recommendations, commenting that
Consumer scoring is not inherently evil. When properly used, consumer scoring offers benefits to users of the scores and, in some cases, to consumers as well. Some uses are neutral with respect to consumers. Consumer scores can also be used in ways that are unfair or discriminatory. The goal of these recommendations is to protect the benefits of consumer scoring, guarantee consumer rights, and prevent consumer harms.
• No secret consumer scores. No secret factors in consumer scores. Anyone who develops or uses a consumer score must make the score name, its purpose, its scale, and the interpretation of the meaning of the scale public. All factors used in a consumer score must also be public, along with the nature and source of all information used in the score.
• The creator of a consumer score should state the purpose, composition, and uses of a consumer in a public way that makes the creator subject to Section 5 of the Federal Trade Commission Act. Section 5 prohibits unfair or deceptive trade practices, and the FTC can take legal action against those who engage in unfair or deceptive activities.
• Any consumer who is the subject of a consumer score should have the right to see his or her score and to ask for a correction of the score and of the information used in the score.
• There are so many consumer scores in existence that consumers should have access to their scores at no cost in the same way that the law mandates credit reports be available at no cost, as mandated by Congress. Otherwise, if a consumer had to pay only one dollar for each meaningful score, a family could easily spend hundreds or thousands of dollars to see the scores of all family members.
• Those who create or use consumer scores must be able to show that the scores are not and cannot be used in a way that supports invidious discrimination prohibited by law.
• Those who create or use scores may only use information collected by fair and lawful means. Information used in consumer scores must be appropriately accurate, complete, and timely for the purpose.
• Anyone using a consumer score in a way that adversely affects an individual’s employment, credit, insurance, or any significant marketplace opportunity must affirmatively inform the individual about the score, how it is used, how to learn more about the score, and how to exercise any rights that the individual has.
• A consumer score creator has a legitimate interest in the confidentiality of some aspects of its methodology. However, that interest does not outweigh requirements to comply with legal standards or with the need to protect consumer privacy and due process interests. All relevant interests must be balanced in ways that are fair to users and subjects of consumer scoring.
• The FTC should continue to examine consumer scores and most especially should collect and make public more facts about consumer scoring. The FTC should establish (or require the scoring industry to establish) a mandatory public registry of consumer scores because secret consumer scoring is inherently an unfair and deceptive trade practice that harms consumers.
• The FTC should investigate the use of health information in consumer scoring and issue a report with appropriate legislative recommendations.
• The FTC should investigate the use of statistical scoring methods and expand public debate on the proprietary and legality of these methods as applied to consumers.
• The Consumer Financial Protection Bureau should examine use of consumer scoring for any eligibility (including identity verification and authentication) purpose or any financial purpose. CFPB should cast a particular eye on risk scoring that evades or appears to evade the restrictions of the FCRA and on the use and misuse of fraud scores. If existing lines allow unfair or discriminatory scoring without effective consumer rights, the CFPB should change the FCRA regulations or propose new legislation.
• The CFPB should investigate the selling of consumer scores to consumers and determine if the scores sold are in actual use, if the representations to consumers are accurate, and if the sales should be regulated so that consumers do not spend money buying worthless scores or scores that they have no opportunity to change in a timely or meaningful way.
• Because good predictions require good data, the CFPB and FTC should examine the quality of data factors used in scores developed for financial decisioning and other decisioning, including fraud and identity scores. In particular, the use of observational social media data as factors in decisioning or predictive products should be specifically examined.
• The use of consumer scores by any level of government, and especially by any agency using scores for a law enforcement purpose, should only occur after complete public disclosure, appropriate hearings, and robust public debate. A government does not have a commercial interest in scoring methodology, and it cannot use any consumer score that is not fully transparent or that does not include a full range of Fair Information Practices. Government should not use any commercial consumer score that is not fully transparent and that does not provide consumers with a full range of Fair Information Practices.
• Victims of identity theft may be at particular risk for harm because of inaccurate consumer scores. This is a deeply under-researched area. The FTC should study this aspect of consumer scoring and try to identify others who may be victimized by inaccurate consumer scoring.

Welfare Surveillance

'Welfare Surveillance, Income Management and New Paternalism in Australia' by Mike Dee in (2013) 11(3) Surveillance & Society 272-286 comments
on the situation of income support claimants in Australia, constructed as faulty citizens and flawed welfare subjects. Many are on the receiving end of complex, multi-layered forms of surveillance aimed at securing socially responsible and compliant behaviours. In Australia, as in other Western countries, neoliberal economic regimes with their harsh and often repressive treatment of welfare recipients operate in tandem with a burgeoning and costly arsenal of CCTV and other surveillance and governance assemblages. Through a program of ‘Income Management’, initially targeting (mainly) Indigenous welfare recipients in Australia’s Northern Territory, the BasicsCard (administered by Centrelink, on behalf of the Australian Federal Government’s Department of Human Services) is one example of this welfare surveillance. The scheme operates by ‘quarantining’ a percentage of a claimant’s welfare entitlements to be spent by way of the BasicsCard on ‘approved’ items only. The BasicsCard scheme raises significant questions about whether it is possible to encourage people to take responsibility for themselves if they no longer have real control over the most important aspects of their lives. Some Indigenous communities have resisted the BasicsCard, criticising it because the imposition of income management leads to a loss of trust, dignity, and individual agency. Further, income management of individuals by the welfare state contradicts the purported aim that they become less ‘welfare dependent’ and more ‘self-reliant’. In highlighting issues around compulsory income management this paper makes a contribution to the largely under discussed area of income management and welfare surveillance, with its propensity for function creep, garnering large volumes of data on BasicsCard user’s approved (and declined) purchasing decisions, complete with dates, amounts, times and locations.
Dee states that
In Australia, reframing welfare governance systems and employment promotion appeals to a rejuvenated neoliberal and paternalistic conception of welfare (Lantz and Marston 2012). At the core of this rationality is a focus on bureaucratic, measurable, rational-technocratic procedures and interventions to ensure compliance and to move welfare recipients into job-search training and employment. Welfare surveillance technologies and investigation strategies are at the heart of this policy suite, expanding the ways the state creates ‘deviants’ out of those who fail to be ‘good market citizens’ and reliable consumers of products and services (Monahan 2008; Maki 2011). 
This article considers some of the implications of the expansion of welfare surveillance in Australia. It begins by briefly outlining the backdrop to neoliberal inspired welfare-to-work policies (a feature of both Labor and Conservative Federal Governments in the last thirty years) concerned with restricting access to welfare in general and, specifically, income support payments. Analysis will be extended to include a discussion of more recent so called welfare ‘reforms’, in the context of ‘activation’ and surveillance. 
The focus is on the BasicsCard, as Australia is the first country in the world to enact this system of income management on identified income support claimants, introduced under the then Howard (Conservative) Federal Government in 2007 (an election year) in the Northern Territory Emergency Response (NTER). It is worth noting that in Australia’s Federal system of Government there are six States; Tasmania, New South Wales, Queensland, Victoria, South Australia and Western Australia, and two mainland territories often treated as states, the Australian Capital Territory and the Northern Territory. There are distinct and at times overlapping responsibilities between the States and Federal or Commonwealth Government, resulting in tensions and disputes over funding and other issues (Healey 2008). For the purposes of this article, the Federal Government only is discussed as it is responsible, through the Department of Human Services, for income support and other social security benefits Australia-wide and is hereafter referred to as the government.
The NTER was and remains, a controversial melange of interventions only implemented in the Northern Territory (NT), ostensibly to address issues raised in the 2007 Little Children are Sacred report, documenting the historic practice of child removal and its negative and on-going outcomes for Indigenous Australian individuals and communities, in the form of child sexual abuse and neglect, domestic violence and a complex array of other problems (Wild and Anderson 2007). Some of the so called ‘emergency measures’ carried out by the Howard Government—which could not bring itself to formally apologise for the forced removal of children featured in the report, while the newly elected Rudd Labor government of 2007 did so—included suspending the Racial Discrimination Act (1975) in order to specifically target Indigenous communities for compulsory health checks and the closure of community employment programs (Cox 2011). 
Income management as one element of the NTER operates by ‘quarantining’ a proportion of a person’s social security payments (including Newstart Allowance, Youth Allowance, Parenting Payment, Sickness Benefit and Special Benefit) to a specialised income management account accessed through the (electronic debit) BasicsCard (Australian Council of Social Service 2010; Billings 2011; Mendes 2012). Income managed monies can only be used to purchase items deemed to be ‘essential’ such as food, fuel, clothing, cleaning products and to pay rent. Prohibited items include alcohol, pornography, tobacco and tobacco products, gambling products/services and home-brew kits, but not chocolate, sweets or other ‘unhealthy’ food items (Department of Human Services 2013). In this way, the BasicsCard can be considered as a paternalistic control/caring, monitoring and surveillance assemblage deployed to secure socially and morally ‘responsible’ behaviours from some of the most disadvantaged individuals and the communities they live in, often denied the services and facilities available in most urban centres. The consequences of compulsory income management for welfare recipients include a sense of stigma and feelings of not being trusted, multilayered complexity, and a loss of control and active decision-making over personal and family finances (Cox 2011). …
[S]ystems of surveillance continue to be expanded, and performance measurement and assessment strategies, investigation and surveillance practices reinforce a deficit-oriented framework focused on perceived individual or family failings (Maki 2011). The commodification of data is now a substantial industry, highly influential in persuading governments of all political hues of the alleged benefits of dataveillance in most policy areas. It is clear that the intense shift to a more technological surveillance oriented system deserves closer attention, particularly in the ways apparently rational andideologically ‘neutral’ systems of surveillance mesh so adroitly with punitive welfare regimes (Wrennall 2010). 
The surveillance of people on social welfare benefits and services funded or provided through the welfare state is far more intrusive than for those on fiscal welfare, with benefits received through taxation and occupational welfare, with benefits gained through employment (Henman and Marston 2008). Moreover, targeting personal behaviour by withholding income payments is inconsistent with a rights-based approach to income support that had been a feature of welfare policy and ideas about social citizenship in Australia since World War II, with the introduction of unemployment benefits by the Curtin (Labor) Government in 1944, ending the practice of applying for State Government ‘sustenance’ in food and blankets, from police stations (ACOSS 2010). This post war settlement around employment and welfare policies became increasingly brittle following the demise of the Whitlam Labor Government in 1975 and throughout successive Conservative and Labor administrations (Kennedy 1982; Pusey 1991; Bessant et al. 2006). Given the pervasiveness of all forms of surveillance and that ‘cyberspace is not a flat, multilateral plane and the nodes of the network are not all equal’ (Fitzpatrick 2000: 390) arguably, the classic formulation by Marshall and Bottomore (1950) of social, political and civil citizenship rights now needs to take into account a form of virtual citizenship rights—or rights over one’s virtual, data self, analogous to rights and protections concomitant with being a physically constituted being (Dornan and Hudson 2003). 
Australia is still, at the time of writing, the only country in the world to impose this type of income management on some of its poorest citizens (Thomas and Buckmaster 2010). While the high tide of welfare surveillance is seemingly irresistible—perhaps even leading to a national identity card—resistance is necessary and powerful and has been an aspect of welfare in Australia since European settlement, with recipients attempting to navigate and subvert an often labyrinthine and administratively perverse system in order to survive (Kennedy 1982; Bryson 1992). The Australian Council of Social Services is calling for the abolition of compulsory income management and the savings made directed to community development projects and greater funding of mental health and other support services (ACOSS 2013). Rollback the Intervention is an Indigenous protest group also resisting compulsory income management and other features of the NTER. In this context, both Indigenous and non-Indigenous people form a growing ‘precariat’ whose circumstances and prospects are entirely tenuous but whose rising anger at massive global and local inequalities holds the potential for a progressive agenda for widespread social change (Standing 2011, 2012). 
The imposition of the BasicsCard is likely to have perverse effects. It is a clear example of a system implemented with little adequate research related to the effectiveness or otherwise of income management, despite governments consistently emphasising the importance of evidence-based policy making (Cox 2011; Buckmaster and Ey 2012). Administrative (and other) costs are high and the question persists as to whether it is possible to encourage or ‘induce’ people to become self-reliant if they are not trusted to exercise real control over vital aspects of their lives. Spending up to $7,900 to income manage people subsisting on $35 a day is highly discriminatory and counter-productive social policy (ACOSS 2013). Flexible and genuinely voluntary forms of income management, alongside increases in benefit rates and the provision of and investment in, a range of other community based support services would be far more effective in addressing poverty and rebuilding trust between all levels of government and citizens. Randa Kattan, Executive Director of Arab Council Australia wryly suggests that, ‘From the bush to Bankstown, people do not need Income Management. They need job opportunities, higher incomes and improved social services’ (ACOSS 2013: 2). So-called ‘welfare reform’ policies that are hastily conceived and rolled out without mindful and adequate attention to the difficulties of local implementation are most likely to fail, being experienced by welfare recipients as oppressive and vindictive (Priest and Cox 2010; Mendes 2012; Karvelas 2013).

24 December 2014

Medicines and Devices Regulation Review

The 110 page Review of Medicines and Medical Devices Regulation Discussion Paper noted here features a large number of questions in eight chapters, with a chapter on Complementary Medicine issues to be released in future.

Public submissions in response to the paper are due 5 January, with a report to the Government by 31 March.

The paper is strongly predicated on cutting what is described as "red tape" and might be construed as exacerbating the regulatory incapacity - if not regulatory capture - criticised by observers of the Therapeutic Goods Administration (TGA), ie the Australian counterpart of the FDA and the EMA.

The paper follows a succession of inquiries into the TGA and into associated matters such as pharmaceutical patents and ACCC recognition of the Medicines Australia Code, along with controversy over fraud and other problems regarding generic medicines.

Chapter Four of the paper states that "opportunities for improvement" reflect five themes:
1. There is duplication of regulatory processes, which creates an unnecessary additional burden on industry.
2. The regulatory framework lacks the necessary flexibility required to facilitate early access to innovative products.
3. Some regulatory requirements are not considered to be commensurate with the risk posed by the regulated products.
4. The regulatory framework is overly complex and poorly understood by many of those who have to interact with it.
5. Some TGA processes are considered to be overly burdensome and out of step with technology. 
Those themes determine the questions asked by the Review panel, which are as follows -

The Regulator
  • How might a trusted overseas regulator be defined? Most countries assess medicines for safety and efficacy prior to approving them for market, but the nature and extent of the assessment that is undertaken can vary markedly. 
  • If Australia were to register medicines based on an assessment by an overseas regulator, how is the Australian public to be reassured that this assessment will have been conducted with the same skill, depth of knowledge, and rigour as currently occurs in Australia? One option may be to develop a set of transparent criteria against which overseas regulators would be assessed in order to designate them as ‘trusted’. 
  • What options are available for determining ‘trusted overseas regulators’? 
  • If a criteria based approach were to be adopted, what criteria should apply in determining whether or not an overseas regulator is trusted? 
  • If the TGA receives an application for registration of an NCE in Australia and the NCE has been approved by one trusted overseas regulator but rejected by another, should the submission be assessed by the TGA? If not, why not? 
  • What other options would ensure that the health and safety of Australian consumers is protected? 
  •  If a trusted overseas regulator rejects an application for marketing of a medicine for the same indications for which that medicine has been registered in Australia, should this spark a review by the TGA? 
Additional Requirements?
  • Should the TGA approve the registration of a medicine on the ARTG on the basis that it has been approved for the same indications by a trusted overseas regulator? If not, why not? 
  • What value do you believe an assessment by the TGA adds in cases where such an assessment has already been undertaken by a trusted overseas regulator? 
  • Are there aspects of safety, quality or efficacy that need to be considered in the Australian context? If so, what aspects? 
  • Would consideration of these aspects necessitate a full assessment of the entire application by the TGA? If so, why? 
  • Should sponsors of medicines that have been approved by a trusted overseas regulator have to submit an Australian specific Module 1 of the CTD to the TGA for assessment? If not, why not? 
  • What do you see as the risks and benefits of not requiring sponsors to submit an Australian specific Module 1? 
  • What would be the likely quantum of savings to industry per application, if Australia was to accept assessments of trusted overseas regulators, with or without a requirement to submit an Australian specific Module 1 of the CTD to the TGA for assessment? 
  • Would such an approach: Result in delayed access to new medicines by the Australian public? If not, why not? If yes, are there strategies that could be put in place to prevent this from occurring? Undermine TGA PBAC parallel processing mechanisms? If so, how might this be managed? 
Meaning of Product Approval
  • Should a change to a medicine that has been approved by a trusted overseas regulator necessitate a further assessment by the TGA in circumstances where that change may impact safety, quality or efficacy? If not, why not? If yes, should the assessment by the TGA be limited only to those aspects of the application that are impacted by the change? 
  • If Australia was to accept approvals of medicines by trusted overseas regulators, should this include conditional/provisional approvals? If not, why not? If yes, should the marketing conditions/provisions imposed by the trusted overseas regulator also apply in Australia? If not, why not? 
  • Should there be capacity for Australia to impose its own conditions, either in addition to, or in place of, those imposed by the trusted overseas regulator and if so, why? Should Australia introduce an accelerated approval program(s)? 
  • What are the potential risks and benefits of such programs and how might the risks be managed and the benefits maximised? 
  • If Australia were to introduce an accelerated approval program: Should there be a single pathway (as per the EU model) or multiple pathways (as per the US approach) to apply? What eligibility criteria should apply to the pathway(s)? That is, under what circumstances could a sponsor apply for accelerated approval of an NCE? 
  • If medicines were to be provisionally approved, based on more limited clinical data than is traditionally required for a full approval: What additional requirements, if any, might be appropriate to alert prescribers and/or consumers to the provisional approval and its implications? What requirements would need to be in place to manage withdrawal of the medicine from the Australian market if safety or efficacy concerns emerged? 
Approval of variations
  • Should Australia adopt a risk based regime for variations, which allows notifications and/or annual reporting for changes that are at low risk of impacting the quality, safety or efficacy of the product? If not, why not? If yes, what might such a regime look like? 
  • How might notification/reporting procedures be designed so as to minimise burden on sponsors? How might the process for minor variations for export only medicines be streamlined so as to facilitate more timely access to export opportunities without compromising health and safety? 
  • Is the Special Access Scheme efficient and effective for Category A patients? Are there issues or concerns with the way in which the Scheme currently runs?
  • Should the Special Access Scheme be revised to narrow the range of circumstances in which TGA approval is required for use of an unregistered medicine in a Category B patient? If yes, what criteria might be applied to determine when an approval is required? If no, why not? 
  • What do you perceive as the risks of such an approach? 
Inadequate post-market surveillance?
  • Does Australia’s post market surveillance of medicines need to be enhanced? If so, how might this occur? 
  • What would be the features of an effective post market surveillance system? If not, why not? Why do you consider the current system effective? 
Regulatory advice to foster innovation
  • Is there a role for the TGA in providing a regulatory advice service to product developers/sponsors? If yes, what should the nature and scope of this advice service be? 
  • How could risks of regulatory capture be avoided? If not, why not?
  •  Is current guidance material easy to locate, navigate and understand? If not, what are the main issues and concerns? 
  • How might this material be improved? 
  • Is the TGA website easy to navigate? If not, how might it be improved? 
Burdensome processes
  • What TGA processes do you consider most burdensome and why? How might these be improved? 
  • Do current regulatory requirements, costs, and timeframes act as a disincentive to the registration of additional indications for medicines? 
  •  If yes, how might the regulatory framework or processes be changed to reduce the disincentives and/or provide incentives for the registration of additional indications, especially in paediatric populations? 
Trusted overseas regulators for generic medicines
  • What options are available for determining ‘trusted overseas regulators’? 
  • If a criteria based approach were to be adopted, what criteria should apply in determining whether or not an overseas regulator is trusted?
  •  If the TGA receives an application for registration of a generic medicine or biosimilar in Australia which has been approved by one trusted overseas regulator but rejected by another, should the submission be appropriately assessed by the TGA? If not, why not? 
  • What other options would ensure that the health and safety of Australian consumers is protected? 
  • Should a change to a biosimilar that has been approved by a trusted overseas regulator necessitate a further assessment by the TGA in circumstances where that change may impact the comparability of the biosimilar? If not, why not? If yes, should the assessment by the TGA be limited only to those aspects of the application that are impacted by the change? 
Additional requirements re generics
  • Should the TGA approve the registration of a generic medicine on the ARTG on the basis that: a trusted overseas regulator has assessed it as being bioequivalent to a reference product available in the overseas market; and the sponsor provides evidence that the overseas reference product and the Australian reference product are identical or interchangeable? If not, why not? 
  • What value do you believe an assessment of bioequivalence by the TGA would add in such circumstances? Could work sharing initiatives between regulators improve timeframes for approval of generic medicines? If so, how? 
  • In what ways could TGA processes for assessment of generic medicines be better harmonised with international practice? 
  • Should the TGA approve the registration of a biosimilar on the ARTG on the basis that: a trusted overseas regulator has assessed it as being biosimilar to a reference product available in the overseas market; and the sponsor provides evidence that the overseas reference product and the Australian reference product are identical? If not, why not? 
  • What value do you believe an assessment of biosimilarity by the TGA would add in such circumstances? 
  •  Should Australia adopt a risk based regime for variations, which allows notifications and/or annual reporting for changes that are at low risk of impacting the quality, safety or efficacy of the product? If not, why not? If yes, what might such a regime look like? 
  • How might notification/reporting procedures be designed so as to minimise burden on sponsors? 
Variations for Export
  •  How might the process for minor variations for export only medicines be streamlined so as to facilitate more timely access to export opportunities without compromising health and safety? 
Burdens on registration of generics
  •  Do you consider the timeframes for assessing applications for registration of new generic medicines to be unreasonable? 
  • How do they compare with timeframes achieved by overseas regulators such as the FDA and the EMA? What do you think are the main factors impacting TGA timeframes for assessment of generic medicines and how might these factors be addressed? 
  • Would the creation of a separate registration pathway for generic medicines achieve more timely approvals and more appropriately align regulatory processes with risk? If not, why not? If yes, what would the generic approvals pathway look like? How would it differ from the process for NCEs? 
  • Should an update to the indications for a generic medicine which is in line with the originator medicine, be treated as a variation, rather than as creating a separate and distinct good? 
  • What are the risks and benefits of this approach? 
Generics trade names
  •  Do you believe that the process for applying for an additional trade name is burdensome? If yes, how do you propose it could be modified? 
  • What do you think would be a reasonable statutory assessment timeframe for consideration of an application for an additional trade name? 
Risks re OTC medicines
  • Do Australian decisions regarding the scheduling and/or rescheduling of medicines appropriately balance risk and benefit? If not, why not? 
  • Are the current scheduling classifications and factors suitable for appropriately assessing substances based on risk? If not, in what way could they be improved? 
  • What would be the advantages/disadvantages of adopting a formal methodology for assessment of risks and benefits to inform scheduling decisions? 
  • What might such a methodology look like? How could the transparency of the scheduling process be improved? 
  • What are the risks and benefits of allowing direct to consumer advertising of Schedule 3 Medicines? 
  • How might any risks be managed? 
Risks re therapeutic goods
  •  Is the threshold at which some products are classified as medicines too low? If so, what criteria could be used to determine which types of products were more suited to regulation as consumer goods? 
  • What are the advantages and disadvantages of excluding some low risk products from the Therapeutic Goods Act 1989
Variations re OTC medicines
  • Should Australia adopt a risk based regime for variations, which allows notifications and/or annual reporting for changes that are at low risk of impacting the quality, safety or efficacy of the product? If not, why not? If yes, what might such a regime look like? 
  • How might notification/reporting procedures and mechanisms be designed so as to minimise burden on sponsors? 
  •  How might the process for minor variations for export only medicines be streamlined so as to facilitate more timely access to export opportunities without compromising health and safety? 
Burdens re OTC Trade names
  •  Do you believe that the process for applying for an additional trade name is burdensome? If yes, how do you propose it could be modified? 
  • What do you think would be a reasonable statutory assessment timeframe for consideration of an application for an additional trade name? 
 Scheduling burdens
  • Are scheduling and registration processes poorly aligned? If so, what approach could be adopted to achieve better harmonisation? 
  • Would efficiency be improved and complexity reduced by introducing parallel processing of: scheduling and registration applications?; and rescheduling applications and related labelling alterations? 
  • What are the advantages/disadvantages of such an approach? 
Recognition of overseas device regulators?
  • Should the TGA undertake its own assessment of the competency of EU notified bodies? If yes, how might this occur? If not, why not? 
  • Alternatively, given the concerns with the EU system, should Australia look to recognise other international regulators as ‘trusted’ for the purpose of device approvals? If yes, what criteria should apply in determining whether or not an overseas regulator is trusted? 
  • Should any criteria take into account different device classifications? For example, a regulator could be designated trusted for some classes of devices but not others. 
Additional requirements re devices?
  • Should the TGA approve the inclusion of a medical device on the ARTG on the basis that it has been approved for the same purpose by a ‘trusted’ overseas regulator?  If yes: ­ should this occur regardless of the class of the device? 
  • ­ How could concerns about the quality of some overseas conformity assessments be managed? If not, why not? 
  • What value do you believe an assessment by the TGA adds? Are there aspects of safety, quality or efficacy that need to be considered in the Australian context? If so, what aspects? 
  • Where there are differences in device classification between Australia and the EU, should sponsors be required to meet additional conformity assessment requirements? If not, why not? 
  • Should Australia adopt the EU classification system? If not, why not? 
  • What are the strengths of the Australian device classification system that cannot be found in the EU system? 
  • Should Australia maintain Australian specific requirements with respect to labelling and post market monitoring? If not, why not? If yes, what value do these requirements add? 
Device approval
  • Should a difference in a medical device that has been approved by a trusted overseas regulator necessitate a further assessment by the TGA in circumstances where that difference may impact safety, quality or performance? If not, why not? If yes, should the assessment by the TGA be limited only to those aspects of the application that are impacted by the difference? 
  • Would this approach apply to all classes of medical devices? 
  • If Australia was to accept approvals of medical devices by trusted overseas regulators, should this include conditional/provisional approvals? If not, why not? 
  • Would this approach apply to all classes of medical devices? If yes, should the marketing conditions/provisions imposed by the trusted overseas regulator also apply in Australia? If not, why not? 
  • Should there be capacity for Australia to impose its own conditions, either in addition to, or in place of, those imposed by the trusted overseas regulator and if so, why? 
  • Should Australia introduce an accelerated approval program(s) for higher risk medical devices? If yes:  What eligibility criteria should apply to the accelerated approval pathway? That is, under what circumstances could a sponsor apply for accelerated approval of a device?  What are the potential risks and benefits of such programs and how might the risks be managed and the benefits maximised? 
  •  If higher risk medical devices were to be provisionally approved, based on more limited clinical data than is traditionally required for a full approval:  What additional requirements, if any, might be appropriate to alert clinicians and/or consumers to the provisional approval and its implications? 
  • What requirements would need to be in place to manage withdrawal of the device from the Australian market if safety or efficacy concerns emerged?
  • What additional post market surveillance would need to be in place for medical devices that were provisionally approved? 
Flexibility
  •  Is the current regulatory framework and classification system flexible enough to accommodate new and emerging medical device technologies? If not, why not? How could it be improved? 
Device risk and regulatory burdens
  • Does the current regulatory framework for medical devices in Australia provide an appropriate balance between managing risk and minimising unnecessary regulatory burden? If not, why not?
  • Should low risk medical devices that are not subject to an independent conformity assessment be included on the ARTG? If not, why not? 
  • Are there any risks involved in not including such products on the ARTG? If yes, why? 
  • What are the benefits of these products being included on the ARTG? 
  • Should Australia adopt a risk based regime for variations, which allows notifications and/or annual reporting for changes to medical devices that are at low risk of impacting the quality, safety, or performance of the device? If yes, what might such a regime look like? 
  • How might notification/reporting procedures be designed so as to minimise burden on sponsors? If not, why not? 
Post-market surveillance
  • Does Australian have the balance right between pre market and post market regulation of medical devices? If not, why not? How could it be improved?
  •  What are the features of an effective post market surveillance system? 
Access to unapproved devices
  •  Is the Special Access Scheme efficient and effective for Category A patients? Are there issues or concerns with the way in which the Scheme currently runs? 
  •  Should the Special Access Category B and the Authorised Prescriber schemes be revised to narrow the range of circumstances in which TGA approval is required for use of an unregistered medical device?  If yes, what criteria might be applied to determine when an approval is required? If no, why not? 
  • What do you perceive as the risks of such an approach? 
In vitro diagnostics
  • Has the regulatory framework for IVD’s resulted in a reduced emphasis on clinical best practice? If so, how.
  • Should there be statutory timeframes for assessment of applications for inclusion of an IVD on the ARTG? 
Process burdens
  • What TGA processes do you consider most burdensome and why? How might these be improved? 
  •  How might the processes required to include a device family on the ARTG be streamlined without undermining public health and safety? 
  •  Are there other concerns with the inclusion of devices on the ARTG? How might these be addressed? 
  • Should the TGA allow a broader range of permissible formats for instructions for the use of medical devices? If not, why not? 
  • Do current regulatory requirements, costs, and timeframes act as a disincentive to the registration of additional intended purposes for medical devices? If yes, how might the regulatory framework or processes be changed to reduce the disincentives and/or provide incentives for the registration of additional intended purposes? 
Classification of devices
  •  Is the classification system for medical devices too complex? If yes, how might it be simplified without impacting public health and safety? 
  • Do manufacturers require assistance, such as online decision tools, to assist them to correctly classify medical devices? If not, why not? If yes, what sorts of assistance would be most effective? 
  •  Is the pre market assessment of medical devices considered overly complex in other ways? If yes, in what way? 
  • What are the major pressure points and how might these be addressed? Is there a role for the TGA in providing a regulatory advice service to product developers/manufacturers/sponsors? If not, why not? If yes, what should the nature and scope of this advice service be? 
  • How could risks of regulatory capture be avoided? 
  • Is current guidance material easy to locate, navigate and understand? If not, what are the main issues and concerns? How might this material be improved? 
  • Is the TGA website easy to navigate? If not, how might it be improved? 
Transparency
  •  Should information about regulatory decisions in respect of medical devices be publicly available? For example, an evaluation report or other relevant information. If not, why not? 
  • What do you see as the risks? If yes, how would this benefit consumers, clinicians and industry? 
  • How could any risks be managed? 
  • Should other regulatory findings relating to medical devices be made public, for example, reports on audits or post market reviews?  If not, why not? What do you see as the risks?  If yes, how would this benefit consumers, clinicians and industry? How could any risks be managed? 
  • Could the regulation of medical devices be made more transparent in other ways? If so how, and what would be the risks and benefits of the proposed approach? 
Interaction with other frameworks (eg the GTR)
  •  Is the system overly complex for manufacturers/sponsors of devices using hybrid/convergent/ co dependent technologies? If yes, how could the process be streamlined without undermining public health and safety? 
Consumer understanding
  • Is the regulation of medical devices transparent enough in terms of informing health professionals and consumers about the level of scrutiny that a device has undergone? If not, how could it be improved? 
  • Should there be a system for medical devices similar to the AUST R and AUST L system for medicines? If not, why not? 
Advertising Framework
  • Should Australia allow advertising of prescription medicines to the general public? If not, why not? If yes, what risks might this create and how could these be mitigated? 
  • What are the risks and benefits of allowing direct to consumer advertising of Schedule 3 medicines? How might any risks be managed? 
Co-regulation
  •  Should Australia continue to require compulsory pre vetting of medicines advertised direct to consumers or should it move towards a self regulatory, or combined statutory and self regulatory, model? 
  • If Australia was to adopt a self regulatory model or a model which combined risk based regulation with self regulation (such as the UK) what key elements would need to be in place to ensure that public health and safety was protected, while minimising regulatory burden? 
Advertising enforcement
  • Should there be a single authority for receiving complaints about the advertising and marketing of therapeutic products? If yes, which agency would be best placed to act in this capacity? 
  • Does the current regulatory framework for advertising medicines direct to consumers: Provide adequate incentives to promote compliance with requirements? If not, why not. What additional incentives should be included? Contain adequate sanctions and penalties for non compliance? If not, what additional sanctions and penalties should be available? 
  • Should the TGA have the power to take immediate action to suspend an advertisement that it considers places at risk public health and safety? 
  •  If yes, why and how might any risks be mitigated? If not, why not? 
  • Is the current self regulatory scheme for advertising of medical devices effective? If not, why not? Please provide examples of where the system has failed.