Welcome to the Progressive Corporation's Investor Relations Conference Call. This conference call is also available via an audio webcast. [Operator Instructions] In addition, this conference is being recorded at the request of Progressive. If you have any objections, you may disconnect at this time. The company will not make detailed comments in addition to those provided in its quarterly report on Form 10-Q, annual report to its shareholders and letter to shareholders, which have been posted to the company's website, and will use this conference call to respond to questions. Acting as moderator for the call will be Jeff Basch. At this time, I'll turn the call over to Mr. Basch.

Good morning. Welcome to Progressive's conference call. Participating on today's call are Glenn Renwick, our CEO; and Brian Domeck, our CFO. Also on the line is Bill Cody, our Chief Investment Officer. The call is scheduled to last about an hour. As always, our discussions on this call may include forward-looking statements. These forward-looking statements are based on management's current expectations and are subject to many risks and uncertainties that could cause actual events and results to differ materially from those discussed during this call. Additional information concerning those risks and uncertainties is available in our 2012 annual report on Form 10-K and our quarterly reports on Form 10-Q issued during 2013, where you'll find discussions of the risk factors affecting our businesses, Safe Harbor statements relating to forward-looking statements and other discussions of the risks, uncertainties and other challenges we face. Each of these documents can be found via the Investors page of our website, progressive.com. We are now ready to take our first question.

[Operator Instructions] Our first question today comes from Vinay Misquith of Evercore.

On your shareholder letter, it seems that management is now, in the second half, more focused on growth in PIF. So curious as to whether that's based on rate reductions that you're taking right now, or is it focused more on the competitor rate actions?

Good, Vinay. That might be one of the comments that's deserving of some additional color, so good use of our time. Let me see if I can get to a lot of the points in there and -- aggregate this year, let's start with that. Aggregate this year, we've actually got rates up very, very slightly. So our aggregate rate change is single-digit basis points positive. As best as we can determine, the marketplace is also relatively low but still positive. I'm doing this off of sort of some graphical interpretations here. I'd say 25 to 50 basis points. So let's set that as a backdrop that, in fact, rates are still going up but relatively minor. As having taken them up last year, even less so, but still positive overall. Last year, we used -- and I think I've talked about this a lot, but I like to sort of get the context and then go on to some of the comments with rate reduction, which, I think, is probably on your mind. Last year, we used what I'll, for this discussion, a relatively blunt instrument. And we took rates up. We call that base rate changes. When we take base rate changes, it's actually uncommon for us to do that. Normally, we're a little more surgical. What we're moving to now and we have been doing since the end of the first quarter -- so this is not actually a new thing, it started end of the first quarter and through the second quarter -- our product managers, this is essentially what they do, is always looking at their product to see different combinations of opportunity for growth at acceptable margins. And in many cases, after using a blunt instrument, we're going to go back and continuously refine…

Yes, that's very helpful. Just a follow-up on that. Curious about what the elasticity is. Historically, I've thought about, when peers are raising rates and if your rates are flat to maybe modestly down, that the elasticity is more and people would shop and you would gain share. So given this environment of roughly flattish pricing, even by peers, maybe modestly up, what elasticity are you seeing, at least in the early phases, for a modest rate reductions on a surgical basis?

Fair question. And again, if we try to do that at the aggregate level, it's not going to be very meaningful. 4 basis points here, 5 basis points there, not going to shift the needle. These rate revisions that we're talking about now are more surgical and can be at levels where the elasticity is significant. In fact, there's no point in doing them unless there is some degree of belief that you're going to get a disproportionate growth for the adjustment that you make. Well, there may be sales where we're simply not getting anything, and that's a pretty simple one to do. But there are ones where we might want to increase the growth. And certainly, you're not willing to do a rate reduction where you get nothing for it. Otherwise, it's just an experiment in margin reduction. So while I haven't answered your question specifically with elasticity, there are sales where the elasticity is actually very extreme. And specifically, in the agency channel, where we have comparative rating, that elasticity becomes immediately apparent in terms of -- think Google Search, where you're positioned in the results return. For our direct channel, it is much more observable in the conversion rate by that individual sector of individuals. So we wouldn't be doing it unless we felt there was a real return.

This is Brian. Just to add on to Glenn's comments about the drivers of unit growth, certainly, the question -- most of the response related to rate levels. This would be more so for the direct channel. One of the things that -- not only did we raise rates in the second half of last year, we also did decrease our ad spend in the second half of last year over that which we had planned. And certainly, given where we are today and in anticipation in terms of ad spend, on a year-over-year ad spend level, the rest of the year should be fairly robust. And so that will also be, hopefully, a generation of increased demand. In fact, so far, in the first 6 months of the year, the demand and quoting activity in the direct channel has been very good. And we are hopeful that with a little bit of an increase in ad spend year-over-year, that will also show quote growth year-over-year.

Next, we have Michael Nannizzi of Goldman Sachs.

Just picking up on that point, I guess, it sounds like in the direct channel, you synchronize the advertising work -- somewhat synchronize advertising and your rate activity. Just trying to understand that a little bit more. I mean, is that what you did last year? And given that you're advertising more now, are you adjusting or do you plan to adjust your rate activity accordingly to maximize the value proposition to customers?

Let me take it slightly differently, that. First, for us, it's a 96% -- grow as fast as possible at a 96%. So always keep that in your back of your mind that we will do that. Last year, approximately this time, the 96% was under pressure. So at that point, we had to get rate at a level that we felt comfortable -- forward rate that we felt comfortable we would want to sell policies at. There is really no great reason to go out and sell a lot of business at a rate level you're not comfortable with. So that was the comment that Brian was making that not only, at about midyear last year, did we do a unusually wholesale review of rates and take up base rates, we also reduced our advertising because those base rates need time to sort of work their way in through the different states that we take them. And so we reduced our advertising until such time as we felt -- and I say reduced. It was certainly not a "Let's drop it off completely," so you saw plenty of advertising. Now we're at a rate level we're very comfortable with. And what I have really just said in the last 5 minutes or so is that we are back to doing what we do all the time. This is the kind of fine-tuning product management, looking for segmentation that can produce a little bit more so I get a little bit more out of this piece of the business. Given that we're comfortable in the aggregate and we'll continuously fine-tune at the micro level, there's no reason for us to be backing off our advertising levels that we had previously planned. But what you will see for the second half of the year is advertising at approximately the same level as you've seen for the first half of the year, whereas last year, you saw a level in the first half of the year and a decline in the second half.

I see, great. And then just one other one, I guess. Given your -- I mean, your size, I mean, you're clearly very large. You're advertising significantly. I mean, your PIF growth has come in, and a lot of that seems like was in response to rate actions that you wanted to take. Is double-digit PIF growth in the direct channel, is that a goal, a target? Is that something that you think that you can feasibly attain? Or from your standpoint, is it more about keeping people in the system than aiming at some potential growth metric down the road?

I'm going to be a little bit hesitant to give sort of guidance there. That won't come as any surprise to regular listeners. But it's -- you mentioned something as if it's just new generation. Keeping people in the system is also a way of growing your PIFs. So every time we have someone leave the book, having to replace that with new, it's a whole lot better if you're keeping the person in the book and getting the new. So it actually is both sides of the equation. And I will just say I do not rule out double-digit PIF growth in the direct channel, nor do I forecast it, but that's what we're about. And I don't want to sound like a broken record here, but that's exactly what we do with the fine-tuning that I just went through. And I would tell you, from my perspective, we're actually -- I feel a lot better about the health of the company right now than at this time last year, where you're having to use those blunt instruments. There's a lot of -- not so much unintended consequences, but there are secondary consequences. Right now, we feel very good. We'd like to sort of hunt for a little bit of growth. That makes perfect sense, but it will never be at the cost of significant margin reduction.

Paul Newsome, Sandler O'Neill. J. Paul Newsome - Sandler O'Neill + Partners, L.P., Research Division: I wanted to ask about your perspective of how the independent channel and the direct channel are linked from a competitive perspective and whether or not that has changed in recent years. If I or anyone else is to take a perspective of an increase or decrease in the general level of competition within one of those channels, should we also be thinking about that naturally bleeding over directly into the other channel? And how do you think about that?

Could you just give me that second piece of that again, Paul? J. Paul Newsome - Sandler O'Neill + Partners, L.P., Research Division: So if I think that there's going to be a lot more competition, say, in the independent agent channel, do I -- should I be thinking of that as also really affecting the direct channel? And has that changed much over the last several years?

All right. Let me see if I can get a couple of pieces of that, and then if I miss them, jump back at me. Directionally, and I can -- I'm only speaking for Progressive here, not a comment on the channel. We have gone, over the last many years, to great lengths to try to get an equalization of the acquisition costs. So we are now distributing in our agency channel at an acquisition cost that is very directly comparable to the costs that we incur in the direct channel. That's a very important statement, and it may not seem it, but it means that there's no real macro arbitrage between one channel or another. And we're being very, very consistent in our outlook to say that consumers will shop how they choose to shop. And while it's very easy for some people to come to a conclusion that clearly, there'll be a massive directional shift one way or the other, that is not supported by the facts. It'll, in fact, be a very slow change, and we've positioned our company so that we are an absolutely equal provider of product to 2 channels without creating an internal arbitrage. There is clearly a difference in price at a unit level between direct and agency. That, we've been very clear on. Agents know it. We know it. Customers know it. But at the macro level, it actually is a similar cost of acquisition. With the competitive structure in the agency channel, we've talked about this 100 times, being really an auction environment with the comparative raters, we feel very good about that. But it also -- and we highlighted this at the IR meeting. We said the cost structure of a company to really compete well on the IR channel was critical. So we, a, feel like we compete well in the independent agency channel. And the number of competitors there is certainly not increasing dramatically. The carriers you know mostly are the same sort of usual suspect group. And in some cases, we feel very good about our cost position there. But we've also done it so that we're not arbitraging against or for our direct channel. So what will happen in terms of channel population will be very much reflective of how consumers want to shop with us. Having said that, I'm not sure if I got at the essence of your question. J. Paul Newsome - Sandler O'Neill + Partners, L.P., Research Division: I think so. I mean, at least from your perspective. There's a lot of concern, I think, that we're going to see more price competition, and Travelers has made this comment, in the independent channel. And I think we're also thinking, do we think about it as sort of a general market issue or a specific-within-segment issue.

Yes, I certainly can't -- or wouldn't talk to another company's actions. But the results are fairly clear, if you take a look over a reasonable period of time, that a cost structure and segmentation structure is very important. And I say both of those and don't weight one over the other because in an auction environment, it's really important to have your rates right at the individual risk level, not at the macro level. And the cost structure is more of a macro issue. We feel very good about our strategy there, and I think it's being supported. I am clearly hopeful that we'll start to be reporting growth numbers in our agency channel. You saw a narrowing of the gap in the second quarter and a growth into the direct channel. I reported at the IR meeting that, in fact, demand in both channels is strong. So we're about capitalizing on that. And the relative competition in the agency channel is not something that I think is higher or different now and don't project it to be significantly different going forward. I am sure everybody is trying their best to do what they do. And with the agent -- with the direct channel, we have, on previous occasions, tried to give you a perspective or our perspective. I'm sure you have one for yourself. But cost of entry for direct carriers is just extraordinarily high, so the likelihood of the landscape on direct changing dramatically seems low to me. And those usual suspects that you know well now will be the ones that will play in that segment. J. Paul Newsome - Sandler O'Neill + Partners, L.P., Research Division: I'd like to -- one more question, but changing gears a little bit. I know you folks have looked at the new accounting ED, and I wanted to know if you have any thoughts and comments and whether or not that might have any effect on kind of how you think about your financial targets.

Brian, you want to take that one?

Sure, yes. We certainly have kept abreast of disclosures. We actually think today's accounting treatment for insurance contracts and loss reserves and the like actually work pretty well. So we're comfortable with the accounting treatments today. And in large part, we have some concerns that the changes will create a little bit more subjectivity and less comparability across companies. So for those reasons, we have a little bit of a concern with them, but we continue to follow it. We continue to participate in discussions about those. I certainly won't issue a comment on the disclosure draft, but that's our current thinking. We think today's accounting for both short-term insurance contracts and loss reserves is pretty good. It's easy to compare across companies, and we think it's adequate disclosures.

Mike Zaremski, Credit Suisse. Michael Zaremski - Crédit Suisse AG, Research Division: First question is on comparative raters. A couple of competitors have talked about the third-party comparative raters within the agency channel causing disruption. I noticed in the 10-Q, you guys talked about helping your quoting activity due to, I guess, California adopting comparative raters. So I was curious, are the comparative raters a good thing for Progressive, and is this -- are the comparative raters only in a select number of states and kind of rolling out into more states as time goes on?

If you'd asked that question about 4 years ago, I probably would have gone into some philosophical rant about whether I thought it was good or not. But now, I would say they're here. That is the landscape, so it doesn't really matter whether it's good or bad. What it means is you have to understand that's how you're going to play in the agency channel. And that puts a premium on the companies who can truly segment and have every price presented. You do not want to win when you have that one bad sell, it will be -- a bad sell, i.e. we price 16-year-olds at the wrong price. So just to make this very simple, that's not quite that simple, but if a company has 16-year-olds at the wrong price, lo and behold, they'll win on the comparative rater on a regular and frequent basis, and you will see that show up in margin. So it puts a premium on being able to have what I'm -- I'm looking at Progressive entirely. So you have views on other companies, what is a fundamental strength of Progressive of segmenting and trying to get every combination of rates right because it will be exposed faster than ever before through comparative raters. One could say that, that means that the loss ratio for the entire industry will be at the lowest level that is acceptable for the industry because there is no opportunity -- or there are lower opportunities for, "I put business with someone, it was slightly higher rate, maybe it was a commission issue." Might have been whatever combination that someone wants, but on a comparative rater, you're going to actually drive to at least a lower level of pricing. It doesn't mean an agent -- you'd have…

No, I'd say no on that. If you call GEICO, you're going to get GEICO's rates. If you call Progressive, you're going to get Progressive's rate. Obviously, Progressive has an asterisk there because we use comparative rating as a way of getting people interested in our company by saying, "These are the kinds of services that you should attribute as brand attributes to us." But the number of people who are proactively going through the comparative rating as part of their direct buying process has actually come down over the years. I interpret that, this is my interpretation, that people feel very comfortable that we're a high-integrity company and a company they like to do business with. But oftentimes, they're not availing themselves of the comparisons that we would be willing to provide to them. They are choosing a company, and I just use Progressive, GEICO, and when they call or more likely go online or go from mobile, they're really only getting one company's rate. What probably, in that case, both of us are doing are making sure that the customer has the opportunity to evaluate alternatives of coverage, of deductibles, those sorts of things, in a very comfortable way, something that agents may have done for them in the past, less so comparatives. Michael Zaremski - Crédit Suisse AG, Research Division: Got it. And lastly then, last question on the year-to-date adverse development from the 2012 action year. How do we think about that relative to the strategy of not increasing pricing that much on an aggregate basis? I guess, does that imply the adverse development is coming from some isolated states?

All right. You're going to get that one? Do you want to take the lead on this?

Well, first of all, auto. The unfavorable development is primarily in the agency channel. Direct is actually slightly favorable through June. And yes, it would be specific states. Not all the states are showing unfavorable development. It would be done on a state-by-state basis. And certainly, as product managers are assessing current rate level as of today, they have that unfavorable development in their loss costs that they build up to figure out additional premiums or what the right premium rate level is. So all that is, as the unfavorable development has occurred, is already embedded into the loss cost that people are using for indications of what rate adequacy is. So I think it's already factored into certain rate level selections, at least in recent months. And the second quarter unfavorable development was approximately 16 million unfavorable. So most of the unfavorable development on a year-to-date basis occurred in the first quarter. I would flip a little bit to commercial auto, where we have had a fair amount of unfavorable development, at least as a percentage of our premium, significantly more there than in personal auto. And there, large component piece of it is in our Truck segment, which is an area where we wrote a lot of business and grew quite rapidly in late 2011, 2012 time frame but have raised rates fairly significantly in the last, call it, 9 months or so. So we believe we've addressed a large part of that unfavorable development already.

Meyer Shields, KBW. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: Glenn, you've talked in the past -- and I'm paraphrasing; if I get it wrong, please correct me -- about how, when Progressive raises rates ahead of its competitors and then they play catch-up, there's a great growth opportunity there. It sounds like that's not the market environment we're in now. And I was wondering if you could compare maybe growth prospects in that scenario against the growth prospects where Progressive is identifying profitable segments where it's under-penetrated.

Yes. No, I think you got the paraphrase close enough to accurate. And when we took rate, let's just go back to the most recent period we took rate, we definitely had an outlook that was, severity-wise, stronger than has materialized. Still rather do that. If I have to make that mistake 10 times out of 10, I'll do it because we have the opportunity to adjust. The alternate is not great. So the fact that perhaps that not materializing as strongly didn't necessarily force the market to react as much as we have seen in the past. So that said, I think it's an accurate paraphrase. Now if you perhaps remember from the IR meeting, I took you through sort of a little bit of a life cycle of a rate revision and how it sort of goes from being slightly overpriced at the point of actually making the rate revision, perhaps in a hypothetical sense, correctly priced at the mid-level of the midpoint of a rate revision, and at the very end of a rate revision, it's now needing rate. The trends sort of extending a little -- or the trend softening means that, that rate, the life of the rate revision may in fact last a little longer. And again, I'm doing this at a macro level, having said this is really tens of thousands of the ordered pairs in the organization. So what we have now is the opportunity to say, "All right, it hasn't materialized quite as strongly." That means we should be able to maintain our rates considerably longer. And good for consumers, and more importantly, we've got a longer period where we can go into the products and start looking for these sales. And I'm not exaggerating. There's tens of thousands of sales. It's not so much growth and margin that everybody is looking at, but it's relative loss ratio between these segments, which is exactly the way we do it. That's powerful. And I think we'll find most of the growth is the fact that we'll be able to continue rate revision, combined with this more surgical approach. So I think you asked me to sort of break it out between the 2. I don't know that I can do that with any precision. But both of those will be vectors that I would be looking for, for our continued health and growth going forward. And to be honest, it's probably 3 months longer than I might have liked. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: 3 months longer of better severity than you considered?

Yes. If the 4 to 6 had actually materialized, then I believe that we might have seen more market reaction 3 or 4 months ago, which would change the competitive positioning. The fact that, that hasn't happened will change our own competitive position by the fine-tuning that I'm talking about, but we'll also be able to keep our rate levels longer than we think we might have been able to. And hopefully, during that period of time, we'll continue to see, as I referenced earlier, still some slight change in a positive direction of the competitors around us. Really, Paul, what I'm saying is we're growing into our rates. And that's okay, too. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: No, I think I understand that. I want to delve into the concept of ordered pairs that you're discussing today because that seems to be a little bit different from growing as fast as possible with a 96%. If, simplistically, if you could go from 93% to 96% and take growth from 7 to 7.1, it might not be a good tradeoff, but it's more consistent with the macro goal.

Very good point. Let me not play more than I would like to play, other than I'm glad you got the point of the ordered pairs. An ordered pair of growing very quickly, pick your number, at a 96% is an acceptable outcome for me. What is less acceptable is a very poor tradeoff from where you are to where you might go. So I think you used an example of about a 10 basis points of growth for a reduction in margin. That doesn't seem like would fit on my indifference curve, if that gives you enough information. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: It doesn't, but I'm probably not going to get more.

Well, let me be clear. If we are at a place -- let's say, we're at a 92% and we're getting 5% growth. Maybe there's elasticity in that sale. But to give up 3 points of margin to get 5.2% growth doesn't seem like a very good tradeoff. And an indifference curve, if you like, would be where would that indifference be? Would it be a 9% growth to give up the margin? Or would you be as comfortable with the 92% and a 5% -- an indifferent point might be a 94% and a 9%. But the example you gave, I said, I don't think would fit on my indifference curve.

Josh Stirling, Sanford Bernstein. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: So to ask a couple of questions on Snapshot, if I may. The USAA news -- congratulations, it strikes me as a big deal. This is a large leading company and an innovator in the business. I expect peers will see this as a pretty meaningful endorsement on the importance of telematics and your IT. And I was wondering if we can get an update on licensing talks. Are you still negotiating with a number of folks? Should we expect more signings? Are competitors going to license the -- actually license the Xirgo devices? And I guess, sort of finally, related to this question, have the terms of your offer changed since the June 30 deadline?

Josh, I hate to -- I usually like to try to give straight answers to straight questions, but let's just say the discussions are ongoing because, frankly, they are ongoing. And in some cases, we're not looking to disclose the names of our partner or the people we're discussing with. So yes on that. I think the USAA, I think you've characterized that very well. Have the terms changed? No. And in fact, we actually closed the period for applying for a license at the end of June, I believe it was. If I'm wrong on that, someone correct me. So we have a good number of companies that are still in flight, in discussions. But I said before, I'm sort of an advocate for level playing field. So we play by those rules as well. We've put our terms, and those are the terms. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: Got it. So shifting gears from the legal, this is obviously a transformational story, but it's been, at least from an outsider's perspective, relatively slow to play out. And it's hard for us to gauge consumer interest in Snapshot and using telematics devices, but the data we have, it still seems pretty low. And I'm wondering, you cycle through some new ads. You seem to be running Parking Lot a lot. And you've got your new iPhone app in test. And I'm wondering if you're finding a messaging and media that works and will actually lead to incremental customers and when, perhaps, we should expect you guys to shift more of your advertising spend to really focus on driving Snapshot.

Yes, very fair. I will tell you that getting consumers to engage in a product that, for the most part, they were never asked to engage in -- and what I mean by engagement is literally getting something in the mail, plugging it in. It's a bigger burden than I think, intellectually, many of us might have assumed. Let me give you some sense of the journey that we've gone on, and I'll keep the numbers because now I see other people, including people that are unlicensed. And therefore, our information will become probably a little more guarded in some cases. But as we started into advertising Snapshot -- let's just go to approximately April of 2011. That's plus or minus; my memory will be close there. We started to actually advertise. And we were getting about, in our direct channel -- and I'll focus on direct right now because it's a cleaner number for us -- about 20% to 22% of our customers taking the option. I've described it as an option because you're going to get a rate. That's going to be the rate you say yes to. Then, would you like the option to in fact put this in and potentially improve your rate? Intellectually, I kind of go, "Why wouldn't 100% of people take that option?" And you know some of the reasons that people are adverse to this, and privacy, and so on and so forth, as well as I do. But the answer was about 22% -- or 20% to 22% of people took it. Then we did some advertising, and we moved that needle to about 30%. And that was roughly where we were last June. We talked for -- in the IR meeting and also a couple of references I've given previously…

[Operator Instructions] Our next question comes from Brian Meredith, UBS.

Glenn, I was hoping you could talk a little bit about the PLEs. And I understood that the rate increases had some impact of that, but is there anything else going on there, any initiatives to try to get that in the right direction?

Yes. Fair enough. The PLEs are just not what we want, probably not what you want. But I'll speak for myself, definitely not what I want. Lots of initiatives inside the company, not going to go laundry list on those, but you can assume that those are going on and would have been going on anyway. So don't assume that's directly related to the rate revision. That's the kind of intensity we've had on retention. There's always some new things. And in fact, I was just reading some of the business reviews on that recently. So happy with the activity there. There's also -- and this was directly related to the more recent environment. There were studies done with regard to both NPS and, to some extent, PLE, which is a more business measure, but NPS is a proxy for that, Net Promoter Score, to see if there are other driving factors other than price. So one of the dangers when you do something like take price up and you see a fall and you say, "Yes, as expected," you certainly don't want to be lulled into some complacency that, in fact, there's also something else going on at the same time and you're not seeing it. So we have looked and looked and looked for that, and we don't see anything that is an additional driver other than price. So that gives rise to my confidence that we will start to see an appropriate reversal of our PLEs to expectancies that feel more consistent with where we were. That will be the price effect. The internal initiatives to try to keep getting PLE extension, those will be on top of that to try and take it to new levels, which is certainly not where we are right now. And I hope -- I doubt that the next conference call will be much more than directional, but -- because these things change slowly, but I hope that, that's a question that we'll be addressing pretty openly 6 months from now with some different answers. But the actions are fine. And at least the primary causation seems to be exactly the way we are reporting it.

Great. And then, the second question. Just curious, I note the policy acquisition ratio is coming down, obviously with the mix shift, as well as some accounting stuff. Is there anything else in there that's been happening with respect to your commission rates or contingencies or anything like that, that's caused it to decline fairly quickly in the last, call it, year? And should we expect it to level out, or should it continue to go down?

I think you meant the expense ratio.

The policy acquisition cost specifically.

Okay. Commission levels really haven't changed much in the last several years. We have previously talked about, I think, of a range 10 to 10.5, and it has basically stayed in that area. So that would not be a driver of an increase in policy acquisition cost. And in fact, I'm actually pleased with our overall expense structure. We've actually made some gains on our expense structure relative to a year ago, and a large part of that certainly is a function of the higher premium per unit. But we've been able to maintain staffing levels at relatively flat staffing levels versus a year ago. So compensation-related costs are growing at less than our premium growth rates. So we're actually making some progress on that. And I mean, we still need to continue to have diligence around that. But that would be core essential to what we do. But the specific question around commission rates changing or contingencies changing much, no, that would not be a driver.

New and real split for direct.

Was that helpful or confusing or...

That's helpful.

Next, we have Robert Glasspiegel, Janney Capital.

I was wondering if I could shift to the upper left side of the balance sheet. You made a successful surgical move into equities a few years back, and you're move into shortening the portfolio and the bonds has taken more patience to play out, perhaps. But with the 70 basis points increase in rates, I was wondering if you could say whether that changes anything dramatically from an investment point of view, recognizing that you tweak within pretty narrow bands, what you might do. But on the margin, is there anything changing?

Bill, do you want to field that one?

Sure. I think on the margin's the right word. As rates grows late in the quarter, and particularly in June, it provided some opportunities to put money to work in the fixed income states with slightly longer durations. We moved the duration out just a little bit, but we're still under 2 years. And then, as rates rose for products other than treasuries, it gave us a few opportunities to find some attractive investments there. But it's not a wholesale strategic shift. We're still and always thinking about our investment portfolio from a total term basis and protecting our capital. So year-to-date, we're still up and contributing to our book value from the investment perspective. And second quarter, we were, call it, either side of flat, which held pretty good, given the environment. But as rates go up, and hopefully more, we'll have a few more opportunities to do something there.

Do you have a specific view on munis, which have gotten hit more than the rest of the bond market this quarter? Some are arguing they that could be attractive, no?

We have added to our muni portfolio. And particularly in June, we were lightened up on it earlier in the year. There, I'd say munis -- longer-maturity munis were hit particularly hard. The front end of the munis held up a fair bit better than the long end did. And we happened to be in the front end of the curve, largely to protect our capital. So our muni portfolio was not hit as badly as longer munis were. And we did add to that somewhat.

So are you saying you're adding munis to the mix, or you're just adding munis within the mix of your normal bond portfolio?

Our muni percentage hasn't changed dramatically, but it's gone up a little bit. We had some opportunities to add munis. They're always part of our mix. So they've always been part of our mix. But there were some attractive opportunities in munis as that market sold off. And there were a fair amount of sellers as you looked at funds flow data. It was across lots of sectors, whether it was high-yield or corporates, munis, et cetera, but particularly -- munis were particularly hard hit from funds flow perspective, so it gave us some opportunity to pick up a few bonds at attractive levels when others were selling more aggressively.

That was our final question. So this does conclude the Progressive Corporation's Investor Relations Conference Call. An instant replay of the call will be available through Friday, August 23, by calling 1 (800) 925-0570 or can be accessed via the Investor Relations section of Progressive's website for the next year. Thank you.