<p>Let's say, in the near future (10 years or so)
I'm not talking about all accountants/CPAs getting replaced, but greatly lessening the need of many jobs if computers can make them much easier, faster, and more efficient. Of course there will always be a human element to the job, but the goal of software is to minimize routine.</p>
<p>What's keeping most CPAs from being worried? Can't complex tax laws be taken into account when designing the software?</p>
<p>Tax laws are pretty complex and have multiple perspectives (loopholes to be exploited and etc). I’m sure that the Big 4 firms have some internal software they use for simple checks but I think most clients prefer to have a human “double check” the numbers to make sure. </p>
<p>Another point could be billable hours.</p>
<p>Software isn’t programmed to stretch the truth is part of it. I’m sure there is some creative accounting on every affluent person’s tax filings.</p>
<p>This is all speculation but a lot of it might be due to the fact that the government requires accounting for all public companies to be done in the same way. I can’t imagine the SEC changing that any sooner than 10 years after a practical piece of software comes out. Because of that, there are few incentives for a company capable of making such software to do so.</p>
<p>In a bit longer time frame, say more like 20-30 years, it’s more than possible that most the non-auditing accounting jobs will get eliminated. </p>
<p>However, you can make the same argument about many engineering jobs. As the world industrializes, there will be fewer and fewer truly unique projects that need to be done in Chemical/Mechanical/Civil fields and, when coupled with better software, I’d bet that a similar percentage of engineers would be lost as Accountants.</p>
<p>My educated guess is a combination of complex tax laws and licensing laws.</p>
<p>I’m sure CPAs use software to do their jobs. If nothing else, you should have somebody who knows something about accounting use complex software systems to do accounting, because that way the software can be used more effectively and obvious errors can be caught.</p>
<p>Also, let’s not forget that there are limits to what software can do. I wouldn’t be surprised if there are problems that arise in accounting that computers can’t do as efficiently as humans, or if there are problems computers can’t do at all.</p>
<p>For instance, I’m surprised that accounting is coming under fire. Why not just automate software construction itself?</p>
<p>At some level of wealth, perhaps the level of wealth that requires a CPA, people expect service. The ability to call someone up on the phone and explain a few things and get an answer or two. Or to meet for lunch to talk about ideas for business or saving taxes or doing a deal. There are also times when you want to have a discussion about things without leaving a record. I’m sure that Steve Ballmer was pretty embarrassed over internal company emails about Vista over the years - things wouldn’t have been so embarrassing if managers had just voiced them in face-to-face meetings.</p>
<p>Apple has been doing well financially in the last few years - they charge double or triple the amount for the hardware that you could buy on a Windows machine. While the perception of better quality, style, design and software are factors in the prices they command, one additional aspect is service. You can bring your Mac into an Apple Store and talk to a real person that will be fairly knowledgeable about your product instead of calling up someone in India, going through voicemail jail and maybe getting an answer.</p>
<p>Because current generations of expert software too transparently ask the question “what do you you want the answer to be?” and then provide a clear forensic link to the resulting answer. </p>
<p>CPAs are much better at ensuring that the question and its answer are both opaque and difficult to link.</p>
<p>Kei</p>
<p>P.S. Same answer for bond rating services, too.</p>
<p>Ok, maybe the more “human element” the CPA/accountant’s job has, the better. But then that would foster the more “sales” type of people to enter the profession, no?</p>
<p>My theory is that, if a job is:
<p>then it can and will be automated. There is no reason not to.</p>
<br>
<br>
<p>I’d suggest reading about the Therac-25 and the conclusions drawn from that case.</p>
<p>One of the big problems with automation is a changing of assumptions that are beyond the parameter of the program. To that end, a clerk or even a child can understand things in the real world where software would have a lot of difficulties.</p>
<p>Ever since our country started industrializing, we have been loosing jobs to new technology. However, the general trend is that technology has freed us from needing to perform repetative, labor intensive tasks so we are able to spend our time doing more important things.</p>
<p>Think about it this way: 100 years ago, the vast majority of Americans weren’t going to college to become architects, doctors, engineers or accountants. Many were working in labor intensive jobs on farms, factories etc.</p>
<p>If you havn’t noticed most of us still have jobs (not taking into account the recession). The trend is that newer technology creates more demand for workers not less. While we loose some jobs and even whole industries to newer technology, these technologies generally provide opportunities for other jobs.</p>
<p>Also, computers cant think. They are very good at doing what they are programmed to do, but at the end of the day, they still need to be programmed by a human. They havn’t figured out a way (yet) to get rid of the human.</p>
<p>ps. I am stil waiting for my flying car that I was promised in the 70’s.</p>
<p>^ If accounting becomes more and more automated, then surely only a few accountants would be needed to do what required many. Then, where would those “excess” accountants go to?</p>
<p>I started working for a manufacturing company after a year in college and I observed the way that they did general ledger. Many people manipulating large tables of numbers on large pieces of paper. The spreadsheet came in and one person could do the work of five or ten. Managing general ledger data arguably doesn’t require special expertise though such expertise might prevent a few errors here and there.</p>
<p>History has shown that those that become obsolete (yes, it’s a harsh word) have to adapt and retrain. This can be very, very hard as someone that has been at the top of their game only to find that they have to restart at the bottom of the heap. Sometimes the timing is very poor in their phase of life. In the past, seeing such changes was harder as information spread very slowly. Today, information spreads very quickly but you have to be watching - from my experience, it is not technically hard to watch but it takes time and effort that could otherwise be spent on other things.</p>
<p>Computer softwares are <em>already</em> available. Turbo tax can help most people do their tax returns; they just have to believe they can learn to use it.</p>
<p>That said, CPAs are needed for complex tax returns and strategy and planning, esp. for corporations. Average joes don’t hire the Big 4 for tax returns. Only the riches with complex income streams and investment schemes would pay expensive fees to the Big 4.</p>
<p>Sam Lee has a great point. I bet companies, organizations and the wealthy will pay top dollar for an accountant to “creatively” massage the books. </p>
<p>Legal of course (or I hope) but I’m sure conventional computer programs won’t have a paper clip popping up and asking “Would you like to make a tax-shelter”?</p>
<p>Long answer approaching.</p>
<p>So, I’m in one of those jobs that “could be replaced by computers.” I thought, as a college student, that I would certainly be replaced by computers. In fact, I thought that I would certainly be typing things into computers all day, that they’d do the heavy lifting, and that I’d get to take the computer’s credit for designing cool skyscrapers and things.</p>
<p>I was wrong. Very, very, very wrong. </p>
<p>It basically falls to this: who would you rather have in charge of an important task?</p>
<p>1) A computer-- The benefits to having a computer be in charge of important things is that it takes a certain input, does stuff to that input, and produces output, and it does it the same way every time. It does this faster than a human is able to, it does it with fewer stupid errors along the way, and it can handle a lot more data more consistently than a human being can.</p>
<p>2) …a human being…-- The benefits to having a human solve a problem is that it takes an independent look at a problem each and every single time. The human makes mistakes, but when the human makes mistakes, it makes them one at a time. Plus, when the mistakes are made, they will either a) be incredibly noticeable to the human (wow, I just calculated that this two-by-four should be able to support a motor torpedo boat… that’s not right), or b) too small in magnitude to be noticed by the human. That means that chances are good that USUALLY the mistakes are too small to be a problem in practice because engineers are required by building codes and engineering ethics to use “factors of safety” to make their designs just a little bit more resistant of the forces it must withstand.</p>
<p>3) …or a human being using a computer? If you have an experienced human being performing that check on each and every single calculation before it goes to final production, then you get the benefit of having things be consistently calculated the same way each time ALONG WITH the benefit of having that experienced set of eyes looking at the output and going, “Yeah, that looks reasonable,” or “No way that’s correct.”</p>
<p>In structural engineering, and I would guess that it’s the same way in ANY calculationally-intensive profession where you could SERIOUSLY screw things up by being wrong, we use the third method.</p>
<p>Computer programs are fickle. I wish I could tell you how often I’m sitting in staff meetings with some of the BEST structural engineers IN THE WORLD (on a daily basis, I drink coffee with the people who’ve designed the majority of the NFL retractable roof stadiums and high-rise hospitals in the United States… sitting around with this group of thirty or so engineers, it’s like, which of our articles was featured in each of the major trade magazines <em>this</em> month?), and someone stands up and says, “Just want to let you all know about a bug in SAP2000 (or RISA, or ETABS, or our own proprietary software that we’ve used for everything for upwards of ten years)-- if you have this and this and this condition, it calculates that there is a zero moment at the end, which means that it’s underreporting the required connection strength for the high-wind condition under certain circumstances. We’ve called our R&D department and they’ve called their contact at C&S, Inc., and they’re working on a patch.” Everyone will shake their heads and look a little nervous, they’ll go back and check everything that they’ve ever worked on that matches those conditions. NOW go read the Therac-25 accounts… And know that the buildings we design aren’t designed for the everyday condition; that’s easy to do. They’re designed for earthquakes, for hurricanes, for high winds, for terrorist attack, and we don’t get to do dry runs or test our proton accelerators on pigs to make sure they work ahead of time.</p>
<p>A couple of other things to consider:</p>
<p>Structural programs are written by either structural engineers (who are lousy programmers and write code with tons of bugs) or computer scientists (who write great code but have no idea what the heck the building codes are talking about).</p>
<p>The only quality control/quality assurance that these programs go through are done by the probably three or four QA/QC engineers that the software company employs, and then it goes out for general usage. You know how you’ll download updates for your phone, or for Windows, or for whatever OS you’re running, or Office, and every now and then, it’ll just crap up your computer and keep crashing? You’ll hear everyone start saying things like, “Hey, don’t download the latest update for the iPhone, mine’s been crashing ever since I did.” The next week, you’ll have to install a patch that fixes the problem, that whoever at Apple or Microsoft has stayed up for five days and nights frantically writing, or something like that. Well, that’s for software that thousands of people use every day, all day long, and that’s after it’s been beta-tested to death and deemed appropriate for release. All industry-specific software is used by… oh, I dunno… a thousand, maybe a couple of thousand users. Less in my field. We’re the beta testers. Not only that, when the software manufacturers release a new version, it’s not that they’ve built a better, shinier mousetrap… It’s that they’ve tried to update the code to match the latest updates from research, from laws, from the tax code, from the ever-changing world of whatever profession they’re writing software for. They’re just trying to keep up. It’s not really V7.0 like the CD case says, it’s V0.1 for the currently accepted practices within the industry. So it’s pretty buggy.</p>
<p>Finally, you’ll notice that you’ve cited CPAs. The “C” stands for “certified”. In structural engineering, we have the PE exam… a grueling exam that lasts for one or two days that you can only take after you’ve acquired four years of experience in your field, and then you can design things on your own. I’m sure that the CPA exam has some similar amount of rigor.</p>
<p>Y’know what certification these programs go through? Nada, zip, zilch. The output they produce is not certified by anyone. It’s not qualified by an accreditation entity. In fact, there are reams of legal documents saying that anything produced by these software suites MUST be used responsibly by a licensed engineer or CPA or whatever, and that the software company isn’t liable for any of that. It’s just a legal nightmare, and the software companies aren’t interested in doing anything other than providing software for a qualified person to use as a part of their professional tool kit. </p>
<p>So the kind of software you’re talking about will never be put together, anyhow. Too risky. Too much liability.</p>
<p>It’s a complex question. I’ve given a long answer. There’s more to it than just this, but Rock Band 2 is calling my name, and I have to not think about engineering too much on the weekends or I go a little bit bananas during the week. Hope this helps.</p>
<p>DH and I have our own small firm, and I use QuickBooks Pro on a daily basis. It’s a great little program, and means that I don’t have to hire a bookkeeper for payroll, monthly taxes, invoicing, etc. But we DO have to hire a CPA a couple of times a year to help us with planning for retirement, annual taxes, HSA’s, etc. There’s no way I could learn all that from QuickBooks. And the rules change CONSTANTLY - it’s hard even for our accountant to keep up with them all. We also like knowing that if the IRS ever audits us, he would go to bat for us, because I sure as heck wouldn’t know what to say!</p>
<p>CPAs must still be in demand, because ours charges a lot more per hour than we can as structural engineers!</p>
<p>
uh…
Innovation(engineering) and routine(accounting) are two completely different things. I don’t see how engineering can be used as an example to show that accounting won’t get replaced.</p>
<p>But like some people mentioned earlier, there are the “human presentation” and “creative scheming” aspects of accounting, which, quite frankly won’t be automated any time soon. But the fact that computers can replace virtually every other aspect of the job makes it liable to massive layoffs in order to increase efficiency.</p>
<p>While accounting might not be as innovative as engineering there can be a great deal of “creativity” depending on the current laws.</p>
<p>See Enron or to a related extent Credit Default Swaps.</p>
<p>^ Yes, but Enron’s “creativity” was against the law, therefore it shouldn’t have ever happened. I would hope that we keep humans in the workforce to enforce the laws, not to break them.</p>