GIS is Dead – Long Live GIS

What is the state of GIS, and where is it going?

Wow, is it even possible to answer that? It seems to be the perpetual question asked at every GIS conference, and embedded into every keynote address given at them. In an attempt to describe the state of GIS, some (many?) are using the terms Paleo and Neo in an attempt to describe past and future tenses of all things geo/gis-related. I’m still not sure these are the correct words to use ( See my blog post titled “My Latest Lesson in GIS”), but they have become a part of the GIS nomenclature, none-the-less.
The impetus to write this post came from a recent discussion on Twitter about how to advise someone just entering the GIS field. It all started innocently enough with a Twitter post by Shawn Bichsel“I’m meeting a potential #geonerd for lunch to spread the GIS gospel. What’s one piece advice you would give someone entering the GIS field” Shawn did a great job summarizing the responses to his question on his blog – Bixel Is Outside.
My response to that question was “New GISers should view GIS as a tool, not a profession. Need to know enviro, transp, engineer, planning, helath, web, etc…”. Then Justin Houk followed up with a tweet “I like your thinking don. You should do a post on your views of GIS as a profession.” And so, here it is.

GIS is a myth

I apologize for the abundance of Twitter quotes in this post, but there were so many great ideas flying around that day, it’s just easier explain my thought process by using them. Bill Dolans said in one of his most profound tweets ever: “IMO, GIS is a myth. There are info systems/apps that use spatial data and methods but GIS doesn’t exist.” This then led to a few more exchanges comparing GIS use and using computers as “text information systems”. And that’s when the light bulb lit up in my head. We are probably at a similar point in GIS evolution as word processing was 30 or 40 years ago.

GIS and Word Processing

People have been drawing maps and communicating through the written word for thousands of years. There was a time when Word Processing was as much a profession as GIS is today. Read this description of word processing from the Wikipedia article, and see if it looks familiar to you GIS folk:

[In the late 1960’s] IBM defined the term in a broad and vague way as “The combination of people, procedures, and equipment which transforms ideas into printed communications,”

Now look at the definition of GIS from the ESRI website:

A geographic information system (GIS) integrates hardware, software, and data for capturing, managing, analyzing, and displaying all forms of geographically referenced information.

Actually, if I had defined GIS that way on a test in my intro to GIS class, without mentioning that a person was needed to operate it, I would have gotten that one wrong.
Anyway, what I’m trying to say is this:

  • There were writers before word processors, and there still are, and always will be, writers.
  • There were geospatial analysts before GIS, and there still are, and always will be, geospatial analysts.
  • Word Processing is no longer a profession, but word processors are still being used as a tool in many other professions.
  • GIS is on it’s way out as a profession, but GIS will continue to be used as a tool in many other professions.

My advice to someone that wants to get into the GIS field is – don’t, because it’s not going to be around much longer. What you should do is – get into a profession you really enjoy, and learn how to apply the various GIS tools to your work.

Reader Comments

  1. Interesting post! I agree that GIS folks must be skilled in another arena. My take is that GIS analysts are inherently similar to statisticians. (In fact some – though not enough – do qualify as statisticians. ) Both apply a certain class of tools and procedures to a variety of problems in wide-ranging fields. As you say, it serves us well to have a deep background in one field and use that expert knowledge to build a project from start to finish. But I expect there will continue to be room for folks to generalize in GIS applications, so long as they understand the spatial realm *and* communicate well enough to bridge gaps in topical specialty.
    GIS proficiency, like statistics, does not come easily to everyone, so there will always be room for the consultant, whether private or public sector. If you’re afraid of commitment but communicate well, you may still have a future in GIS. But it wouldn’t hurt to take some straight stats courses, just in case…

  2. I agree with you, with one major caveat: GIS as software is infinitely more complicated than word processing as software. I believe there will still be a demand for the next 10-20 years for GIS analysts and/or programmers who can do the highly specialized, esoteric aspects of GIS that demand an understanding beyond casual use. And this is going to be the part about GIS that differentiates it from word processing as a career.
    I think on-demand web programs like Google Earth and Bing are going to evolve and become easier to use in such a way that the demand for basic GIS programmer is going to become extinct. But if you develop a background and understanding of how a particular industry operates, and you learn the complicated aspects of GIS that separates the amateurs from the professionals (such as scripting, etc.)… that will be what creates job security for the GIS professional.

  3. Don,
    I’ve been quietly ranting about this for a couple of years now. GIS is the name given to “doing nifty things with spatial data”. For a long time, these nifty things were a black art that required special software, and arcane training. These days are ending.
    As spatial data becomes a first class data type in all databases, we’ll see much broader adoption and use of spatial data, albeit transparent to the end user for the most part. We’ll still have *some* analysts to do ad-hoc stuff, but for the most part, “where-ness” will be just as common, and easy to use as integer and text data types.
    I think we can see this with ESRI releasing tools like MapIt – free of ArcObjects, this tool integrates directly with SQL 2008’s spatial data type. I think it’s still a little “GIS-y” in that it’s ALL ABOUT THE MAP, but it’s a step in the right direction. Over time I think we’ll just see the standardization of the “spatial data” input / display control – just like we have standard text areas, check boxes and drop downs. We’ll squirrel all the fancy overlays/intersects/topology checks away behind the scenes leaving the user blissfully unaware that “GIS” was even in the mix

  4. I have long said “there is no such thing as GIS for the sake of GIS” – it takes a combination of many things to make it successful and effective.
    Knowing how to swing a hammer does not confer knowledge of architecture or structural engineering.
    GIS works hand-in-hand with other disciplines and domains in order to solve problems, just as the hammer works hand-in-hand with other disciplines, e.g. structural, HVAC, electrical, plumbing – to create a building.
    As far as “neo” vs. “paleo” – that’s also long been overblown and overrated discussion fodder as well. Neo kids on the block come along and think they’ve reinvented the world. Yes, the mashups are cool – but once you get beyond the slippy map eye candy, it’s back to the question of “where’s the beef” i.e. analysis. Where IS the analysis? It’s more than just layers and basic visualization. The “paleo” folks have a wide variety of analytical tools and techniques yet unknown to the “neo” crowd.
    Further, you have folks so wrapped up around web APIs and wrinkles that basic concepts of what makes good cartography starts to suffer, as things get juggled for efficient web delivery – not to mention, questions of geodesy and other things starting to dawn on the “neo” folks, which the “paleo” folks have already been dealing with for decades.
    So – some of that “paleo” stuff IS still arcane.
    So what does it all mean, and where do we all go with all of this? Likely things still continue as they have, with various partnerships, between geographers, domain experts in various fields, from law enforcement to archaeology and so on, to computer science folks, to remote sensing folks, even to social media, ambient data and crowdsourcing/mining and to the hackers and next-unrealized bunch of “neo” kids after the current crop, who find yet another set of great ways to exploit technology to further the cause. Still no “GIS for the sake of GIS” – but instead, continued strength through synergies.
    From John Snow’s 1854 London Cholera outbreak visualization (and even prior) to the present, things just keep evolving. Techniques may have changed, but the basic idea has not, it’s just been ever more empowered.

  5. Great discussion. I want to add that this way of thinking suggests that instead of teaching GIS (like we used to teach word processing?), we need to use it across the curriculum (like we do word processing!). The long term future of education in GIS is not teaching GIS, but rather using it across the curriculum. That’s not easy to sell or implement, but I believe it’s the way forward.
    The other thing to bear in mind, I think, is that “desktop GIS” is slowly dying, so learning it is not a long term career move.

    1. Response to Adena Schutzberg’s comment –
      Thanks for the feedback Adena. I’ve seen the recommendation that we change the way GIS is taught mentioned in a few other places in this discussion, and I agree wholeheartedly with that. When I took my first GIS classes in grad school, the only department that offered them was the planning and geography department. That was 8 years ago, and I don’t think that situation has changed much. However, I disagree with the notion that learning desktop GIS is not a long term career move. The growth of GIS is not happening in one direction, it’s happening in all directions. The fact that we see tremendous growth in the online application of GIS does not mean there is movement away from the desktop side of things. Desktop might be growing at a slower rate, (or maybe it’s just under the radar), but growing it is. In fact, my advice to “…get into a profession you really enjoy, and learn how to apply the various GIS tools to your work.” more often than not, will require the use of desktop GIS. There is only so much one can do with online/web-based GIS tools. Complex spatial analysis needs the horsepower that only desktop GIS can provide. Web-based tools will never be able to satisfy all of my needs as a planner.

  6. The twitter side of this yesterday was my best twitter experience so far… maybe I’m finding my flock! 🙂
    I was going to bring up Geographic Information Science, but saw @Taleisn’s tweet of link to Dawn Wright’s article, so that’s covered there.
    Also related in my mind: “education vs. training” conversations. To me, education is about learning to ask good questions, training is about how to use tools to answer given questions, i.e., science vs. tools.
    I particularly agree with Don’s closing recommendation: get into a profession you really enjoy, and learn how to apply the various GIS tools to your work.

  7. Pingback: Spatial Isn’t Special « James Fee
  8. URISA has struggled with an issue tangential to this – but still very closely aligned. Since URISA is an education-based professional society, we try to do our best to offer the best services/networking/educational opportunities/etc to our membership (and of course to attract more members in the process!).
    Way back in the day – before there even was something commonly known as GIS – the Urban and Regional Information Systems Association was formed to provide a place for *technology* as it became more ubiquitous in local and regional government (on those fancy new 8088s). As GIS evolved, URISA was the only game in town for people looking to learn about GIS. But that was 40ish years ago and now there are lots of organizations for geospatial folks to find kinship and continued learning.
    So URISA struggles to figure out how to carve out a niche in a market that didn’t have niches when it first started.
    We started thinking of ourselves at the place for the GIS (or geospatial) Professional.* I really like that as a market niche for URISA – it allows for people who really run GIS applications, say an enterprise system within a larger government IT operation, as members. It also allows for folks like me (and you?) who need to use GIS (with varying degrees of technical skill) in our daily jobs – and it’s vital to that work.
    *I’m not talking about GISP as a certification – that’s a whole different conversation.

  9. IIRC, this discussion is at least 17 years old. I remember back that long, hearing and discussing whether GIS was a discipline unto itself, or a tool used within disciplines. Of course, it was, and is, both.
    I also recall back then that there was much discussion about spatial data development being conducted in an institutional way (SDI now?) or on a project-as-needed basis (Google now?). Still the same discussion.

    1. I think the confusion about whether geography is a factor of all disciplines or a discipline unto itself is at least as old as the earliest ancestor of the word “where”.

  10. My favorite history professor used to begin each class by hanging one of those beautiful old rolled maps (this particular one of Europe) at the front of the room. Every exam he gave had a geographic component and, whenever a student would complain about it, he would say: “History happens some time and some PLACE.” He was a cantankerous old pain in the ass, but he had a point. And I don’t think he even realized that he was ‘doing GIS’.
    I think many of us tend to lose sight of the fact that ‘GIS’ was around long before computers were. For years now, I’ve been pointing to the fact that a phone book is a geographic information system. Think about it – it’s a bunch of data, geographically located. Isn’t that – in a nutshell – what GIS IS?
    I know that phone books lack maps, but I also know that GIS can exist without maps. It doesn’t happen very often, but it does, indeed, happen. And maybe if it happened more often, we’d all be a little less confused about what GIS actually is. The heavy reliance on maps tends to make us think that GIS is geographically determined. It’s not. Rather, it’s geographically INFORMED. There’s a difference, and it’s an important one.

  11. What a fantastic discussion. Due to the number of Certified GIS Professionals, the history of the computer being used for spatially focused purposes and the evidence of a market (jobs, contracts, projects, and academic discourse); I pose the challenge that there is in fact a GIS Profession at this time. The dynamics of society and rapid change could render this to not be true at some time in the future, however, today there seems to be a readily identifiable number of people who see themselves as GIS Professionals.

  12. GIS is a powerful tool for all kinds of occupational specializations but we will always need professionals. Maps are are often thought to reflect objective truth when in actuality, all maps lie, and we have to be careful about where we lie and where we tell the truth.
    I see people using GIS all the time who don’t understand basic geospatial concepts. I also see lots of terrible cartography: un-intuitive symbolization, ineffecient use of page space, cluttered symbolization, and not understanding the limitations of the data they are using, which can lead to unintelligible or deceptive maps. I talk to a lot of people who use GIS but dont understand the basics of database theory.
    Per Terry’s remark about the history professor who did not realize he was using GIS, that’s because he was using GEOGRAPHY. GIS uses automated tools to link data to spatial locations. Geography and cartography have been around a lot longer than GIS and I think we tend to conflate the tool with the underlying science and knowledge base.
    Per the word processing methaphor, we all know a lot more about writing than we think we do. We were taught the skills beginning in kindergarten and they did not stop teaching us until we were done with school. Geography is a second string subject, not taught at most grade levels, rarely taught by professionals and often reduced to memorization of state captials. GIS makes it easy to make a map, but making a good map takes a lot more.

  13. This thread leads Muki Haklay (above) to pose a rhetorical question: what is master’s level education in geographic information science and technology (GIS&T) for? The thread came to my attention while leading a graduate seminar on professionalism in GIS&T. In the seminar I stress that thought leaders should be expected to marshal evidence and solid reasoning when making claims about the professions that occupy the GIS&T field. Don’s claim–that “GIS is on it’s way out as a profession”–is provocative, and usefully so (it has provoked several responses from our class). However, the claim is based neither on evidence nor sound reasoning. No evidence is presented that the number of individuals seeking careers or already employed as GIS professionals is declining, because there is no such evidence. And in place of reasoning there is only a weak analogy to word processing. As Sean Gorman points out, statistics is a more apt comparison, and it is anything but declining as a profession. Muki concludes that the masters degree is an opportunity to “focus on principles and on improving specific skills” beyond the standard GIS driver’s license. The most basic of those principles, perhaps, is that scholarship still counts.
    Thanks for the discussion!

  14. I cant say i agree that GIS is a dying profession but more a profession that continues to develop as do all professions.
    I recall my job 15 years ago to my job now and how much being a GIS professional has changed in those times. I recall using GIS when there wasnt a GUI interface… my how times have changed.
    Just because the tools of a profession are seen to be becomming simpler and facets of GIS are being incorporated into daily lives doesnt mean a profession is dying, nor does it mean that one shouldnt pursue a future with it. Tools of almost all trades develop, become simpler and make their way into the public arena but there will always be a need for those who really know what they are doing. Any one can make a map but that doesnt mean your working in GIS it just means you have the basic skills to use some of the tools GIS professionals use.
    Very interesting to hear different points of view. A great discussion.

  15. In no way is GIS a dying field. I work in Global Census and every govt office is strapped in terms of GIS expertise. I think people that do not understand GIS (unfortunately often statisticians) tend to say it is dying. Ive been in the field for 15 years.
    BTW…I can do any calculation SAS SPSS can do using GIS.
    My 2 cents

    1. “In no way is GIS a dying field.”
      Yeah, it is growing like crazy in India. India must now be ESRIs number one customer. That is, if they are using legal software and not hacked copies.
      The GIS field is dead or dying here in the U.S. I have never seen it so bad in my 23 years in the GIS industry. Pick a job that can’t be sent through a wire or learn to speak Hindu. I cannot recommend a GIS job to get into or a degree in.
      Crazy times we are living in!

  16. So what I’ve gathered is that GIS is too complicated for most people to learn so there’s no future in it. Interesting. Maybe understanding GIS is difficult because everybody thinks they already know what it is. Yes, it IS a bunch of different things and yes, there ARE other ways to implement individual aspects of GIS through other avenues. However, GIS is not ArcMap, it is not GeoMedia, it is not GRASS or another software program; it is an entirely distinct and free-standing discipline that has been unable to be fully accepted as an extremely useful and powerful tool because…everyone obviously knows what it is…except that it’s too complicated to understand unless you’ve been doing it for long enough and have had extensive training…? I don’t understand the details of nuclear physics, so it’s obviously useless.
    There were mathematicians before there was an abacus, there was a slide rule before there were calculators, there were calculators before there were supercomputers able to process mathematical problems and concepts no human could ever do in a lifetime. These computers process enough data in an hour to make someone’s head explode. On this train of thought, computers, or programmers/technicians etc. will no longer exist because we have computers.
    See where this doesn’t track?

  17. While I agree that the GIS profession as a whole is not dying, but rather evolving, it is the way in which it is evolving (or may have already evolved) that is of concern.
    The role of Senior GIS Analyst, or any related title, is seriously threatened by changes in the education system. Knowledge that was once confered only in multi-year programs at a limited number of institutions, is now available as a 15 hour certificate in most major universities, and even as an online continuing education certificate for high school graduates and other professionals at many. The principal theories and analytical processes that have protected the role of the Senior GIS Analyst are now available for the low low price of a few hours and a few hundred dollars.
    Engineers, Landscape Architects, Urban Planners, Transportation Planners, Environmental Scientists – they now have the ability to log on to a certain GIS software company website, and register for any one of a wide variety of GIS training programs designed to instruct the masses – exactly as said GIS software company intended. Couple this with the aforementioned GIS continuing education certificate, and these professionals are now armed with enough basic knowledge to be relevant in the field of GIS. And contrary to what we all want to admit, they’re doing it! What these professionals have, that the straight-up GIS professional does not, is the additional knowledge of their primary career. Not only does an urban planner know how to perform a particular spatial analysis, but he/she knows WHY. Senior GIS Analysts can not claim this.
    Thus, we are left with two clear roles in GIS:
    -The GIS technician, who, by design or lack of professional opportunity, has not developed any programming skills. These sorts are relegated to the mondane day-to-day data management and routine map production tasks that (1) others don’t want, and (2) probably isn’t what drew them to the field of GIS in the first place. The result is rampant job dissatisfaction and high turn-over.
    – The Developer – I purposefully left “GIS” out of the title. The Developer now is someone who, like the Analyst, chose an entirely different career path and has now supplemented their original knowledge with that required to program GIS. This is no different that how developers have kept up with the rapidly evolving development world, so it is foolish to assume GIS is too much for them to learn. Anyone who can learn C++, Java, Javascript, .NET, PHP, ASP, and how to develop in a variety of APIs, can learn ArcObjects, which ever platform flavor. Developers may also be someone who began in GIS, but these types are having to invest an increasing amount of time and effort in development training to acquire all the necessary knowlege on that side. But I haven’t even discussed the rising threat of offshoring development positions!
    So, IMO, if you’re in GIS and don’t plan to acquire additional training outside GIS, be prepared for a career of menial work.

    1. Thanks for the comment, Brian. I like your analysis of the state of the GIS profession, and generally agree with your division of it into two roles. I will, however, note a couple of things from my perspective.
      1- I prefer “GIS User” rather than “GIS Technician”. The use of the word technician implies a limited role of the GIS user in his/her respective main profession. The word user (for me, anyway) implies a role that is incorporated more fully into a larger profession.
      This actually leads to my second note:
      2- I do not agree that ALL GIS Users are ” …relegated to the mundane day-to-day data management and routine map production tasks…”. Granted, I have met many users that fit this description perfectly. However, as the GIS profession evolves and grows, more organizations are recognizing GIS is an integral part of their respective professions. They are incorporating GIS directly into their workflows instead of relegating it to a separate department.
      I am very lucky to have clients that recognize this. My expertise with GIS is usually what gets my foot in the door, but my planning background is also viewed as valuable.

  18. All the people that I know that are left working in GIS are in government jobs with the Feds and Cities. They are now worried that they are going to get let go from their easy and great benefit jobs in the next two years if Obama gets thrown out. VERY few private GIS jobs left for U.S. workers. The private sector pretty much offshored every GIS job they could back in the early 2000s.
    Outsourcing now has killed off ALL growth of GIS job market in the U.S. First, it is an easy job to outsource because it deals with data and not people. For example, city planning jobs deal with people in city government and their community. GIS jobs follow a simple spec that any Tom, Dick, or Apu can do. Second, GIS is fairly easy to figure out. If jobs are going to be cut back the GIS people will get the axe first. The people in planning and other departments can do their GIS work or it will be outsourced. GIS workers and their departments usually fall under these larger departments which will absorb the GIS department in really bad times. And as we all know, it is REALLY bad job times. And if the Republicans win big in 2012 there probably will not be any government GIS left. They will have us using eche-a-sketches to do GIS work.
    People need to step way back and look at where things are headed with GIS. I see more personal tablet and smart phone GIS apps and less large GIS project work.
    I think the monopoly that ESRI has is dying due to the insane amount of money they still charge for the software. Government can no longer shell out $11,000+ for a full ESRI ArcMap license. The gouging the government days are over with big overpriced software contracts. They don’t have the money to pay for highend software anymore! There are free GIS packages out there that work really well like QGIS that can replace a lot of the low end ESRI Arcview licences. I would learn to use as many of these new GIS software packages, since some city and county governments may soon be forced to use them. Also, if Google decides to add free GIS related tools to Google Earth, it will be another nail in good paying GIS jobs and ESRI software.
    As I have told many, it is better to just minor in GIS and learn as many GIS software packages as possible. Things are changing so fast and going more to a personal user level and the cloud.

  19. I will say I was a fool for getting a Geographic Information Science BS degree (degree based in computer science) combined with working as staff in scientific academia right after graduation. Half my BS classes involved learning and applying advanced geospatial concepts traditional statistics, geostatistics, spatial analysis, and intense programming. Here I am doing asset management (primitive data management plan, metadata expert) and technician-level work 100% of the time the past 2.5 years. Every chance I’ve asked to use any of the advanced skills I have had much experience with in my BS career has been shut down by my boss (by silence). Everyone else with an MS that works here essentially gets free training from me (part of my job description) and eventually get to apply these skills on projects because of their MS degree. That is the absolute worst thing I can think imaginable for someone’s career. Not only do I make myself obsolete, but have nothing to show for.
    I will say it feels pretty shitty to have to pay more for an MS degree just to unlock the potential to use existing knowledge/skills. It’s such an artificially-imposed hindrance to natural growth and application of talent not being able to apply knowledge/industry subject matter learned on-the-job. I’m the fool for thinking a degree that pretty much is a means to an end would be more desirable than having a ‘degree’ in the industry of the subject matter itself. #FML

Comments are closed.