The Debate Over the Definition of Basic Income

1 A Note on Linguistics

Language develops by use. No ruling, no official body can make something the “right” or “wrong” definition of a word. There are common and uncommon, standard and nonstandard uses of words. Official definitions can stipulate a term for internal use, but beyond that, they are no more than an effort to influence language in a particular direction. Language might or might not follow.

Basic income is still outside most countries’ political mainstream, and still unfamiliar to many people. The basic income movement has limited influence over the language at this point. Media and social discussions of basic income reveal widely divergent and sometimes conflicting understandings of what it is.

Anyone hoping to affect how language is used should recognize the limits of their influence. Therefore, we need to keep this effort simple and take lessons from what kinds of terms tend to catch on and which don’t.

2 Framing the Issue with a Two-By-Two Matrix

This section examines the interaction between these two questions by plotting the answers on a series of 2 × 2 matrices. Table 1 introduces the matrix. The top row (cells 1 and 3) identifies a yes answer to the question of whether the definition is restricted to a grant large enough to live on (or any other threshold). The bottom row (cells 2 and 4) identifies a no answer, below the threshold. The left column (cells 1 and 2) identifies a yes answer to the uniform question: a non-means-tested grant. The right column (cells 3 and 4) identifies a no answer: a means-tested (i.e. income-tested) grant. Therefore, the four cells are.

Cell 1. Yes, Yes: a uniform payment, above the threshold.

Cell 2. Yes, No: a uniform payment, below the threshold.

Cell 3. No, Yes: a means-tested payment, above the threshold.

Cell 4. No, No: a means-tested payment, too small to live on.

The first concept that needs identification is not an individual cell, but the union of all four: the entire set, the family of conceptions of income guarantee programs that may or may not be livable and/or uniform with respect to income. If we use “basic income” to identify the entire set, we employ a “broad” or “expansive” definition: open to yes or no answers to both questions as shown in yellow in Table 2 below.

Choosing the expansive definition of “basic income” would not get us out of the need for more terms because any discussion of the broad conception naturally brings up the question of what kind of “basic income” is under discussion.

Another potential definition of basic income stipulates that the grant is uniform (with respect to income) but not whether it is large enough to live on. It is the union of cells 1 and 2: the yellow area on the left side of the matrix in Table 3. BIEN has used this concept of basic income since it first voted on a definition of the term in 1986, with this aspect of the definition remaining in place after the 2016 revision.

This definition has made it into some dictionaries. The Oxford English Dictionary (OED) defines “universal basic income” as “Financial support provided by a government in the form of standard, recurring payments to individuals without the need for pre-qualification; … a scheme based on this type of financial support, or an income provided by it.”^[10] The Cambridge Dictionary’s definition is more explicit on this point: “an amount of money that is given regularly to everyone or to every adult in a society by a government or other organization and that is the same for everyone.”^[11] Dictionary.com and Merriam-Webster.com have no entries for universal basic income.^[12]

If a uniform grant is under discussion (whether called basic income or any other name), the contrast between it and an otherwise similar means-tested grant is usually important. If so, we need a term for the green shaded area in Table 3, the union of cells 3 & 4 in Table 1: a means-tested grant whether or not it is large enough to live on.

Another candidate to be the definition of basic income incorporates the threshold requirement but remains neutral on means testing: the yellow-shaded area at the top of Table 4, the union of cells 1 and 3 in Table 1.

If a livable income guarantee is under discussion (by any name), the contrast between it and an otherwise similar but less-than-livable grant is usually important. That concept is shown by the green-shaded area in Table 4 or by the union of cells 2 and 4 in Table 1.

The most restrictive definition of basic income requires a yes answer to both questions: a non-means-tested, livable grant: an unconditional income that is both uniform (with respect to income) and above the threshold. This concept is designated by the yellow-shaded area (cell 1) in Table 5 If this restrictive concept (by any name) is under discussion, it raises three other questions. First, what do you call a grant that is above the threshold but means-tested (Cell 3, shaded green in Table 5)? Second, what do you call a grant that is uniform (with respect to income) but below the threshold (Cell 2, shaded green in Table 5)? Third, what do you call an unconditional grant that is neither livable nor uniform (cell 4, shaded green in Table 5)?.

Adding up the number of concepts mentioned in the discussion of Tables 2-5 gives us nine things that might need names. Table 2 has 1 definition. Table 3 has 2. Table 4 has 2. Table 5 has 4. And 1 + 2+2 + 4 = 9 – nine conceptions of basic income even in a discussion limited to these two questions. Some of these concepts come up a lot – others less often.

Although we do have to decide which of these nine concepts should be called “basic income,” no group declaration will resolve definitional issue(s). We might not need names for all nine of the concepts identified here, but we will only resolve the definitional issue when we have nonpejorative terminology that people on all sides of the relevant debates are willing to use for all the relevant terms in this typology.

With this in mind, the rest of this article discusses options that might help facilitate respectful debate.

3 Using the 2 × 2 Matrix to Understand the Current State of the Definitional Debate

Tables 6–9 show what I believe are the most common terms in use in English now. Terms for the nine concepts exist, but most of them are either controversial or not-common-enough-to-be-considered-standard.

As mentioned above, “basic income” is used in at least four different, partly overlapping, and partly conflicting ways, and each one is a candidate to be “the” definition. The use of most of the other terms is equally inconsistent if more from lack of standard use rather than from controversy about what is best.

The most recognizable name for an unconditional-but-means-tested grant is “the Negative Income Tax” (NIT), which was the standard name for that concept in the 1960s and ’70s when it was a widely discussed policy option in the United States and Canada.^[13]

The term NIT was still common when BIEN held its first conference 1986. Attendees probably thought it obvious that basic income would become the name of the uniform grant while the means-tested grant would continue to be known by its familiar name, NIT.

But in recent years, the term NIT has fallen out of favor. The concept of a “negative tax” was never as easily understandable to laypeople as it was to the economists who coined the term. It connotates negativity. And it is associated with ungenerous proposals put forward by neoliberal economists, such as Milton Friedman.^[14] Most of the progressive supporters of unconditional-but-means-tested grants today want to avoid all these associations, and use of the term NIT is declining. Although both the Cambridge and Oxford dictionaries have entries for it,^[15] if the people who support the concept won’t use the term, NIT’s day as the standard term has already passed. No term to replace NIT has not yet become standard, but this use of the term “guaranteed income” is becoming more common in the United States and to a lesser extent in Canada.

Further complicating the issue of naming this concept is that there is no agreement about the relationship between means-tested and uniform unconditional grants. Some supporters of the means-tested grant like to think of it as practically the same as basic income (under BIEN’s existing definition of the term), and have even used the name “basic income” for it, implicitly or explicitly using the expansive definition of basic income. Because of the efforts of people with this view, means-tested grants have come to dominate the discussion of “basic income” in some countries, such as Canada.^[16]

Many supporters of unconditional uniform grants object to this use of the term, seeing it as an effort by supporters of unconditional-but-means-tested grants to benefit from the positive associations the basic income movement has built up over the last several decades, even though the movement was made up mostly of people pressing for unconditional, uniform grants. This use of terms also tends to marginalize discussion of uniform grants because there is no word to distinguish it from the means-tested version of the grant other than the largely archaic and distinctly unattractive term “demogrant.”

Other supporters of unconditional-but-means-tested grants think that the differences between it and basic income are substantial, even seeing it as a distinct rival of basic income.^[17] Therefore, they believe it is important to distinguish their proposal from basic income. Some of these supporters have started to call this policy the “guaranteed income,” although it is sometimes unclear whether “guaranteed income” is being used specifically for means-tested grants designated by cells 3 and 4 or whether it is being used for the expansive family of ideas designated by the whole 2 × 2 matrix.

Basic income supporters who prefer the uniform model vary in the extent to which they view means-tested grants as a variation on or a rival to what they call “basic income.” But they are overwhelmingly in favor of using the term “basic income” exclusively for the uniform model and in wanting a term that clearly distinguishes between it from the means-tested (i.e. income-tested) version of the unconditional grant.

The threshold issue sparks a somewhat different controversy. Most people within the basic income movement support a livable grant. However, not all of them agree that the definition of “basic income” needs to incorporate a threshold. See below for discussion of that debate.

If livability is left out of the definition of basic income some modifier is needed to distinguish between livable and non-livable levels of basic and/or guaranteed income. The typology above recognizes three modifiers to designate livability: “full,” “livable,” and “sufficient.” The obvious counterparts of these respectively are “partial,” “non-” or “less-than-livable,” and “less-than-” or “in-sufficient”.

The pair of “full” and “partial” is simpler and more intuitive than the other possible pairs of modifiers I’ve suggested to designate livability, but this pair also has baggage.

“Full” has been used not only to designate a livable basic income, but also to designate one high enough to replace all other state provided cash transfers. Any use of “full” and “partial” to designate livability would have to overcome that association. But I don’t think that barrier is insurmountable. The use of “full” and “partial basic income” in the livability sense is already far more common than the other sense. Most countries have at least some livable cash transfers. So, “full” in those two senses might be close to being synonymous.

With these issues in mind, the next three sections discuss typologies based on the expansive definition, BIEN’s definition, and the restricted definition.

4 A Typology Based on the Expansive Definition and the Simple Modifiers: an Elegant but Possibly Unworkable Solution

If we were to use the expansive definition of basic income, only four modifiers would be necessary to identify all 9 terms, but as explained below, I believe, the solution is unworkable. Tables 10-13 display this possibility. These tables use “universal” and/or “unconditional” as the modifier designating uniformity with respect to income and “guaranteed” as the modifier designating variability with respect to income. I discuss difficulties with this typology below.

These tables use “full” and “partial” as the modifiers to distinguish between livable and non-livable grants.

These sets of terms are elegant, understandable, and simple. This strategy is the simplest way to preserve the popular term “basic income” and to identify all 9 terms with only four modifiers.

In fact, if we were starting from scratch, I would probably suggest that this typology as the best solution. But we’re not starting from scratch. Because of the particular history of the use of the term “basic income, this set of terms is likely to face resistance, cause confusion, and fail to catch on. The concept of uniform, unconditional grant we now call “basic income” has been around at least since Thomas Spence’s lecture of 1775.^[18] The term, “basic income” has been used for that concept for nearly a century.^[19] “Basic income” gradually started to become the standard name for that concept in the mid 1980s when organizations like the Basic Income Research Group (UK) and the Basic Income European [later “Earth”] Network (BIEN) were founded. “Basic income,” “universal basic income,” and “unconditional basic income” have been used synonymously at least since the 1950s, perhaps since the 1930s. Any attempt to pry them apart would face resistance and sow confusion.

However, this synonymous use is not well known enough outside the community of scholars and activists working closely on this issue to definitively rule out this possibility. The Cambridge and Oxford dictionaries both define universal basic income as a uniform grant, but neither of them has entries for “basic income” or for “unconditional basic income.”^[20]

The Basic Income Canada Network (BICN) seems to be moving toward something like this typology in its webpage, “Basic Income Explained,”^[21] which discusses the “universal” and “guarantee” models of “basic income” but stops short of naming them.^[22] However, when I have heard BICN members who support the “guarantee model” refer to a uniform grant, they have resorted to using the term “demogrant” rather than “universal basic income.” Canadians who continue to support a uniform unconditional grant and prefer to call it “basic income” or “universal basic income” have complained that they feel marginalized by this turn of events.

A large number of people in the worldwide basic income movement consider uniformity with respect to income to be an essential characteristic of the policy they like to call “basic income.” Many of them also want to limit the association between it and a means-tested grant – even though the means-tested concept shares every other characteristic of “basic income.” Similarly, many means-tested-grant supporters, especially in the United States don’t want their term to be too closely associated with that rival. Thus, the above typology will face resistance from both sides in the debate between the means-tested and uniform version of the unconditional grant.

I believe these problems are insurmountable. This solution will not work despite its elegance and simplicity.

5 A Typology Based on the Current Definition of Basic Income

If the typology in Part A doesn’t work, we cannot both retain use of the popular term “basic income” and identify all nine concepts with as little as one term and four modifiers. Nobody wants to drop the term “basic income.” One possibility is to build a typology on BIEN’s current definition of a uniform (non-means-tested) grant without regard to size.

Tables 14–17 illustrate one such typology, which identifies the 9 relevant concepts with two terms and four modifies.

The dialogue in the United States seems to be moving in this direction with people on both sides of the uniform-versus-means-tested debate converging on the use of “basic income” for the uniform concept and “guaranteed income” for the income-tested concept. One advantage of this typology is the main two terms in it (basic income and guaranteed income) have come to be used in the same way by supporters of either model – and largely by critics of the two models as well. This shared terminology rose gradually from decades of dialogue between supporters of each concept and critics of both.

In the past, “guaranteed income” has been used as an expansive term (including both means-tested and uniform versions), and today, it is often used in ways that make it unclear whether it refers specifically to means-tested grants or to the expansive conception. This problem notwithstanding, “guaranteed income” makes a good contrast with “basic income,” and the use of these two terms in that way provides fairly simple and elegant solution.

Income Movement promotes a uniform grant under the name “basic income” and uses “Basic Income Guarantee” for the expansive concept (although their website does not always make these definitions clear).^[23] The Guaranteed Income Community of Practice and the Economic Security Project promote means-tested grants under the name “guaranteed income” and they both contrast that idea with “basic income” in the uniform sense of the term.^[24]

The U.S. Basic Income Guarantee Network (USBIG) uses “basic income guarantee” for the expansive concept, “basic income” for the uniform concept, and “Negative Income Tax” for the means-tested concept.^[25] USBIG has tried to popularize the term “basic income guarantee” since 1999 but with only limited success.

Nevertheless, these two terms are still a long way from becoming widely recognized as standard. Sometimes experiments are referred to interchangeably as basic income or guarantee income experiments whether they examine a uniform or means-tested grant.

There is no standard term for the expansive concept in the United States (or in the English-speaking world as a whole).^[26] The clearest way I know of to refer to the expansive concept in the United States right now is to say, “basic or guarantee income.” The lack of a single term for the expansive concept might be this typology’s biggest weakness, but the phrase “basic or guaranteed income” is not difficult to say. It has the same number of syllables as the phrase “universal basic income” (eight).

The drawback of having no term for the expansive concept is that the means-tested and uniform versions share one extremely important attribute: they are both unconditional in the sense that they have no behavioral requirement. That characteristic makes them both very different than most social policies in most welfare systems throughout the world. It might be useful to have a term for that class of policies – or maybe it’s easy enough to say, “basic or guaranteed income.” I do not know.

If people are not content to say “basic or guaranteed income,” we would have to either popularize the term “basic income guarantee” or come up with an alternative to it. Existing alternative names for the expansive concept don’t seem much more likely to catch on. Although this article has used “income guarantee” in that sense, given the new-found popularity of “guaranteed income” in the means-tested sense, “income guarantee” is unlikely to become standard. Many other terms are possible, but to introduce one would involve a willingness to start from near-zero recognition. It takes a concerted effort to go from there to common usage.

Other terms in the typology are also far from standard. No term like “full” or “living” has become standard for a livable unconditional grant, but the pairs of “full” and “partial” or “living” and “partial” are intuitively easy to grasp. Most of the recent U.S. experiments have been “partial” in this sense, but few of the researchers and reporters writing about them have felt the need to stress that fact.

6 A Typology Based on the Restricted Definition

Now consider possible terms when “basic income” is used in the restrictive sense. Table 18 shows four fairly straight-forward terms for the four individual cells. “Basic income” in the restrictive sense (both livable and uniform). A “guaranteed income” in a similarly restrictive sense is large enough to live on and variable with respect to income. The less-than-livable versions of these two policies are distinguished with the addition of the word “partial” in front of them.

Table 18 introduces four fairly straightforward, intuitive and easy to understand concepts, but the typology is less straightforward when you consider what to call the other five basic income-related concepts in the full typology, and how to get people who focus their attention those concepts to use terms consistent with the narrow definition of basic income. Consider Tables 19–21.

Table 19 lists some of the many possible terms for the expansive concept. All of them work if people adopt them. But all of them are very far from wide recognition, much less acceptance.

Table 20 reveals what is probably the biggest problem for a typology based on the most restricted conception of “basic income.” What term do you use for the concept matching BIEN’s current definition, which is also the dictionary definition of “universal basic income:” the left column in Table 20 and the union of cells 1 and 2 in Table 3? No term for this concept is obvious when a threshold is introduced into the definition of basic income, and the most intuitive terms, shown in Table 20, aren’t very appealing. A person who wanted to fastidiously protect the definition of “basic income” as a livable, uniform grant would be forced to say something like, “basic income or partial basic income.” The phrase would get tedious with repeated use and might not be easily understood.

A similar problem exists with the terms in Table 21 below.

There are at least eight reasons to be skeptical both of the usefulness of this typology and of the likelihood it will catch on.

First, as mentioned above, the goal of defining the terms in this typology should be to facilitate clear communication for people on all sides of the debate. I’m afraid that a decision to use the restrictive definition would sacrifice this goal in favor of the goal of defining the favored term in such a way that no bad plan could ever fall under the definition of “basic income.”

The desire to incorporate livability into the definition seems to be driven by the fear that people will use the name to promote regressive policies.^[27] Such proposals do come out from time to time. The replacement of existing policies targeted at the poor with a universal and uniform basic income with a high marginal tax rate on low-income people can constitute a regressive – perhaps highly regressive – change if the basic income is too low.

We don’t lack for the words we need to explain that these proposals are bad. They usually come from outside the basic income movement and only rarely use the term “basic income” anyway. It is the sacrifice of other programs and high taxes rates for low-income people that make them regressive, not necessarily the low-level of the basic income that makes the overall plan regressive. If livability is the goal, almost any basic income, no matter how small, is a step in the right direction all else equal.

Not all proposals for a less-than-livable basic income are regressively motivated. Some less-than-livable proposals are seen as transitional policies – the best we can do under current conditions in hope of eventually increasing it to a livable level. Insisting on a restrictive terminology might be a barrier to that strategy of implementation. Other proposals tend to be issue-specific, such as a pollution-tax dividend or a sovereign wealth dividend. It’s useful to have a term that facilitates discussion of the links between these policies and livable basic income proposals.

Second, the restrictive definition of basic income suffers from what we could call “the full professor problem.” There is no official academic rank called “full professor.” Officially, full professors are the only true “professors.” Lower ranking “faculty members” such as “assistant” and “associate professors” aren’t officially “professors.” The phrase “faculty member” has failed to catch on as term for the whole set (and contains ambiguity of its own). So, despite the official definitions, they continue to refer to all “faculty members” as “professors” forcing people to use the unofficial term “full professor” to clarify what they mean when referring to people officially ranked “professor.” This clarification reenforces use of the term “professor” for all “faculty members.”

In the same way, no matter how hard one might try to restrict the name “basic income” to a livable grant, people will often be forced to use some word like “full” or “livable” for clarity, and doing so, they will unwittingly reenforce the perception that the size of the grant is not part of the definition of “basic income.”

People who want to add a threshold to the definition of “basic income” will succeed only if they come up with a term that people who want to talk about the no-threshold concept will willingly adopt. I doubt any such term is available, and therefore, the effort will fail to catch on even if many basic income networks incorporate a threshold into their definition.

Third, the restrictive definition also suffers from what we could call the Trekkie problem, which means that you usually lose when you try to fight the dictionary. By the 1970s, the word “Trekkie” had made it into many dictionaries as the word for “Star Trek fan.” Many dedicated fans didn’t like the term and tried referring to themselves as “Trekkers” instead. A half century later, “Trekkie” remains the only word for “Star Trek fan” commonly found in dictionaries. Fans might have been better off embracing the word and working to develop positive associations with it.

The lesson is don’t fight the dictionary. The battle will be long and difficult. You’ll usually lose. And you don’t win anything terribly significant if you win. Except in extreme cases, you can usually communicate the ideas you want to communicate without changing the definitions already in the dictionary.

The dictionary is your friend. Work with it rather than against it as much as possible. It took decades of work to get “universal basic income” into dictionaries. It will be far more difficult to make a revised use of that phrase so common that dictionaries have to revise their entries. Pick your battles wisely. If you want to talk about an unconditional grant that is large enough to live on, you can make that clear by putting words like “full,” “livable,” or “sufficient” in front of basic income and be perfectly well understood.

Fourth, many of the existing programs and recent experiments that people discussing basic income look to for evidence are less-than-livable grants: the Alaska Dividend, the Namibian study, GiveDirectly, the Indian study, many of the current U.S. projects, and so on. Discussion of programs like these are going to continue to be a major focus of the basic income discussion. What word would we use for them to signal that they’re close to but not quite basic incomes? How will we get people running these programs or advocating programs like them to use that word rather than “basic income?” People will probably continue to call them “basic income,” and at best, we’ll get them to put “small” or “partial” in front of it. Invariably we’ll need to clarify the difference between these programs and livable basic income by putting some word like “full” in front of basic income. We’ll run right back into the full professor problem, reenforcing the current BIEN, Oxford, and Cambridge definitions, and reenforcing the use of modifiers. If we’ll be driven to use modifiers anyway, we might as well embrace them.

Fifth, a threshold-based definition requires a precise identification of that threshold, but monetary thresholds are difficult if not impossible to identify with much precision. How much, exactly, is “substantial” or “enough to live on” or “enough to live in dignity” or to ensure “social inclusion?” All of these criterions imply different thresholds, each of which is vague and at least partly subjective. It’s not easy to say exactly what quantity and quality of goods qualifies as enough for subsistence, much less for dignity, inclusion, or substantiveness. The price of goods varies day-to-day and place-to-place, even neighborhood-to-neighborhood. A threshold-based definition brings in the problem of what to call a grant that’s near the threshold, possibly a little above or below. With a threshold definition, we very often won’t know for sure whether a grant of a given amount is or is not a basic income until precise inflation data is available a year or so later. Suppose the grant is clearly above the threshold in some parts of the country, clearly below it in others, and questionable whether it is above or below it in others. What word, other than basic income, should people use to talk about that grant? And how will people who only want to talk about the restrictive concept get them to use it?

Sixth, a no-threshold definition of basic income facilitates discussion of multiple thresholds: for example, a substantial basic income, a livable basic income, a basic income sufficient to maintain social inclusion, and so on. Basic income defined by one particular threshold inhibits discussion of alternative thresholds. This might be an example of sacrificing the goal of facilitating clear communication for the desire to have a word that can never be associated with a bad plan. Or it might be the attempt to solve a political disagreement (about which threshold is best) by definitional fiat.

Seventh, although one might hope a threshold definition would bring unity to the movement by separating out people who want a regressive change, a threshold definition does not resolve conflicts: it creates new conflicts over which threshold is the right one and whether a particular proposal passes that threshold at any given time. Attempts to enforce uniformity of belief tend to splinter movements.

Eight, along with the difficulty of defining a precise threshold is the issue that the big changes we want to see seldom occur all at once as one crosses a precise threshed. I have written extensively on the need for a basic income large enough to give people power over their lives, independence, the power to say no to bads job and abusive spouses. Such power requires livability and more.^[28] But in those works, I’ve tried to stress that there are levels of basic income that are clearly too low to give someone that power; there are other levels that (given other background conditions) are clearly enough; and a large grey area in between where it gives people some of that power but perhaps not enough of it. I’ve also tried to stress that, given our starting point, a basic income of almost any size is a step in the right direction. Experience in Alaska, Namibia, Kenya, and other places shows that people use what power a very small basic income gives them to take back some of the power over their lives.^[29]

Imagine a policy that is exactly like basic income except that its level is a little too low to give people 100 % of the power we want them to have. It only gives them 60 %, 90 %, or 99.99 % of that power. Is it better to insist that none of these policies can be called basic income at all or to say that these are basic income plans that are a little too low? Obviously, I think the latter is better.

Toru Yamamori asks, “Is a Penny a Month a Basic Income?”^[30] But that’s the wrong question. No one is talking about a basic income of a penny a month because it’s irrelevant. Yamamori doesn’t show any examples people promoting penny-per-month basic income plans, nor is his intent to address the problem that someone might start talking about such a policy. His intent is to rule out all unconditional grants below the livability threshold. In that case, the better question would be: is a penny less than one particular measure of livability a basic income?

Two other good questions are: do we lack the words to clearly explain that a penny a month is a bad plan without changing the current definition of basic income? If not, why go through the difficult and perhaps hopeless effort to get people around the world to change their use of terms? Threshold-based definitions are not worth the trouble.

Yamamori’s article does an excellent job tracing the history of the use of the term “basic income” to show that the people who coined and popularized that term were intending to for the word “basic” to mean enough to cover basic needs.^[31] This was a surprise to me, because I had thought of “basic” in terms of providing a non-zero “base.”

Yamamori does not address the important question of why, despite their intentions, this sense of the term failed to catch on?

I think it failed partly for reasons discussed above and partly for the reasons mentioned in the linguistics note above. The people who attempted to popularize “basic income” as a threshold concept failed to consider that people would need terms for a less-than-basic income and for the set of including both the [full] basic income and less-than-basic income. Thanks in part to the full professor problem, the common usage adopted their phrase, but altered its meaning to fit the needs of the wider dialogue.

Unlike the term, “living wage,” the livability of “basic income” is not self-explanatory nor necessarily clear from context. Without good terms that people would want to use for below-the-threshold or non-threshold incomes, common use of the term “basic income” lost its connection with any threshold at all. Linguistic generalization like this is very common and usually beyond anyone’s control.

It’s not enough to say “I define basic income as this” or even to add “I define this other word for that related concept.” You have to name related concept in ways that appeal to people who regularly talk about those concepts. No such term is forthcoming from a typology based on the restrictive definition, and so it is likely to fail.

7 Discussion and Conclusion

Although I have argued that some typologies are better than others, the main point of this article is neither to suggest a particular definition of “basic income” nor to suggest a particular typology. The main point of this article is to show that the issue is far more complex than the simple question of how to define one term.

The existence of a set of related concepts implies the need for restraint on the part any group that seeks to impose their terminology on others without seeking agreement between people who focus their attention on different concepts in the set.

People will continue to support means-tested and/or less-than-livable grants even if basic income networks leave them out of their definitions. Unless we can find respectful terms that clearly distinguish the relevant concepts and that people on all sides of these debates feel comfortable using, different groups are likely to keep using the term “basic income” in very different ways, and confusion will continue.

So, I suggest the following guidelines for resolving the definitional issue.

First, be aware of the entire typology. If we choose a narrow definition, we need words – that people on all sides of the issue will actually use – not only for what basic income is but also for what it is not. If we choose a broad definition, we need words for different kinds of basic income.

Second, don’t try to resolve every problem by changing the definition. Simpler definitions are better. Change the definition when we lack words to discuss issues not to protect the political appeal of a phrase. Therefore, avoid the impossible desire to come up with a definition that rules out any bad plan from being called “basic income.” The devil is and will always be in the details. There will be good and bad basic income plans. To try to rule out all bad plans by definition would require an ever larger, ever more technical definition that would be harder and harder to understand. And even then, I doubt it would successfully eliminate all possible bad plans from fitting the definition.

This effort does not resolve disagreement; it furthers disagreement. One person’s good plan is another’s bad plan. If we try to resolve all our disagreements in the definition, we won’t get beyond fighting over the definition. We have to leave room for disagreement.

Longer definitions have unintended side effects. The attempt to eliminate type I errors (allowing a bad plan to fall under the definition of basic income) increases the amount of type II errors (excluding a good plan from falling under the definition). For example, one suggestion has been to add the stipulation that a basic income has to be unform not only with respect to income but also relative to all characteristics other than (possibly) age. The intention is to rule out discriminatory basic income programs, such as giving less to women or ethnic out-groups.^[32] But it would have the side effect of ruling out any effort to simplify social policy by combining basic income with need-based support programs. For example, suppose a country introduced what they called a basic income program of $A for normally abled people, $B for blind people, $C for deaf people, $D paraplegics, $E for a group that has been awarded reparations for past discrimination, and so on. Strict application for the uniformity stipulation implies that such a policy is no basic income at all. It might be better to leave this conception of uniformity out of the definition and rely on other protections against discrimination, such as the principle of equal protection of the laws – or simply by saying that discriminatory basic income plans are bad basic income plans.

This discussion suggests that a penny a month is or can be a basic income. Possibly, but because it cannot significantly help maintain real freedom for all;^[33] because it cannot help protect freedom as the power to say no;^[34] because it cannot make a substantial difference in the lives the people who need it most,^[35] it is also a bad basic income plan. A very bad basic income plan. One can effectively oppose this and many other bad plans just as well (and perhaps better) by saying it is a bad basic income plan than by trying to define basic income in such a way that no bad plan will ever fall under the definition of basic income.

Third, when trying to get words into common use, agreement is more effective than imposition. It’s easy for people who support the broad conception to say that people who support the narrow conception should distinguish it from other kinds of “basic income” by calling it a “demogrant,” but it’s difficult if not impossible to get them to play along. Similarly, it’s easy for supporters of a narrow conception to say that people who prefer means-tested grants should call them “minimum income,” but it’s difficult if not impossible to get them to play along.

Accept the limits of your power and influence. People seldom resolve disagreements by definitional fiat. Ask yourself, what are that chances that our proposed term will actually catch on – actually replace the existing dictionary definition? How are you going to get people who disagree with you on policy to use your terminology?

Although basic income is usually defined to be uniform with respect to income, many people prefer the means-tested version of the unconditional grant. They will talk about means-tested grants more than people who support uniform grants. If you want this group to use a term that clearly differentiates these two ideas, you have to find a term that they are willing to use. Simply telling them that basic income is not their word is unlikely to work.

Fourth, pick your battles. Ask yourself whether you want to have a battle with groups that disagree with you on that definitional point. If you consider them opponents, and you think you can win, maybe you should go to battle. But if you’re not sure you can win, or if you consider people who support the related policy to be allies, it might be better to talk it out with them and find terminology both sides can agree on.

Fifth, closely related to picking your battles is the need to recognize when you’re starting one. I’ve mentioned that political disagreements are seldom resolved by definitional fiat. I’ll add now that when they do, it is usually by marginalizing the opposition. We need to ask ourselves whether the people on the other side of the issues defined by these two questions are people we would like to marginalize?

Some supporters of income-tested grants in Canada, for example, have said that they don’t care much about definitions, that anything in the expansive concept is “basic income,” and that the difference between income-tested and uniform grants is mere “detail.” But the only model of “basic income” they want to talk about is the means-tested version.^[36] This strategy has had the effect of marginalizing people who don’t agree that the difference is a mere detail.

Anyone pursuing this marginalization strategy should recognize that they are doing it and that by doing it, they are creating an adversarial relationship with whom they disagree on a mere “detail.” They should consider whether they want to make adversaries out of people who differ on this detail. If not, they should probably talk to those people and find terminology that both sides are happy to use – even if one side is much larger than the other.

Arguably, BIEN could help settle this issue. But I warn against thinking that BIEN or any national or regional group can settle it unilaterally by clarifying its definition. Although BIEN’s definition is regularly cited as an authoritative definition, BIEN’s definition is usually cited by people who like BIEN’s definition. People who want to include means-tested grants or exclude grants below a certain threshold can easily avoid citing BIEN. BIEN is not influential enough that people who prefer to use the concepts excluded by its definition are likely to feel compelled to use it BIEN’s way or to use terms in the typology that goes with BIEN’s definition.

BIEN might be able to be of more help by trying to bring together people with different ideals to discussed terminology hopes of finding one set of terms that they are all willing to use. This sort of meeting would be relatively easy in North America,^[37] where the Basic Income Canada Network, Guaranteed Income Community of Practice, the Economic Security Project, Income Movement, and USBIG focus on different parts of the typology and would benefit from agreeing to terminology. Whether such a meeting would produce a meeting of minds is another question. Such a meeting might be useful in other places as well. It is likely to be more difficult in places where one model or another dominates the discussion.

Despite efforts to harmonize usage, the definitions of “basic income” and related terms are likely to remain controversial for the foreseeable future. That means everyone who discusses the issue should clearly define their terms and be respect of people who uses terms differently.