Greg's Opinion Greg's big blog of whatnot


HISD 2014 Redistricting: Austin HS Hearing

I spent last night embedded at the fourth and final HISD Redistricting field hearing at Austin High School, located near the University of Houston campus. As dumb luck would have it, I had to tend to dad's passing away last week, which contained two redistricting hearings. Among the things you find out when you get back in the saddle of redistricting chatter is that it doesn't take long for these things to get repetitive.

Gene Locke does the standard opening info to a crowd of something less than 10 or so (including myself, excluding legal/translation/cartographic staff). There will be a private session of HISD's board to discuss legal matters involved in redistricting. And the next time we have this conversation in public will be at the Aug 14th public session. Since there was no overly-heated discussion at the North Forest hearing, I'd have to think that the proposed map is looking fairly solid for approval. Again, if you want to rifle through the numbers, the info packet is here.

As better luck would have it, I do have shapefiles for both the "interim" map that incorporated North Forest ISD into HISD trustee districts, as well as the proposed plan being presented at these hearings. If you want to check the 2011-era plan, here's a link to that for your enjoyment. The "interim" plan, for all intents and purposes, is irrelevant for the sake of comparison. The proposed plan is embedded here for your clicking and zooming fun.

» full page
» Google Earth file

I've got a request in for calculating those lovely CVAP numbers as well. I'll update whenever I get those results back.

For the light turnout at last night's affair, the questions were pretty good: one asked about the use of 2010 Census data; another about the fairness of representation for Latinos given the relatively low citizenship rates, and Why there were so few public hearings? All things considered, it didn't sound as if there were significant major questions about the drawing of the districts. So there's a chance that the plan could be approved as-is at the August meeting. Wait and see.


NFISD Total Population Overlay


Why Voting Precinct Detail and Geography Matter

» NY Times: Mistrust in North Carolina Over Plan to Reduce Precincts

My kind of detail ...

When Alan Langley, a Republican member of the local elections board here, explains a new proposal to consolidate five voting precincts into two, it sounds procedural and well-meaning: He speaks of convenient parking and wheelchair access at the proposed polling places, and of saving more than $10,000 per election.

Those precincts, however, are rich with black voters who generally vote Democratic. And when the Rev. Dante Murphy, the president of the Cleveland County N.A.A.C.P. chapter, discusses the plan, he talks of “disenfranchisement” and “conspiracy.”

“We know,” Mr. Murphy said, “that this is part of a bigger trend — a movement to suppress people’s right to vote.”

The bitter disagreement in this city of 20,000 is part of a broader voting rights battle charged by race and partisan politics that is happening in a number of communities, many of them Southern, where changes to election laws no longer require advance approval from the federal government after a year-old Supreme Court ruling voided a key section of the Voting Rights Act.

The Lege here in Texas passed HB1164 to allow for something along these lines in Texas. The stated goal was to reduce some of the "Precinct Inflation" that comes from redistricting. I'm not well enough versed in how it is that the recent round of redistricting bloated the number of precincts from 885 to over 1069, as opposed to the old days when we had over 1200 precincts that wound down to 885 in years past.

The layout of the original bill is below. The author was a freshman and fell for some of the confusion I had in reading the bill: it references "wards". As Ed Johnson (at about 10:55) clears up later in testimony, "ward" just means city council district in statute. Some of the other issues of ADA-compliant polling locations are also mentioned as reasons that the change is needed. And as Johnson goes on to mention, there was an additional bill by Johnson's former boss that would have limited the number of precincts even further.

I haven't seen anything on either the Tax Assessor or County Attorney sites to indicate that there are proposed changes, but I would think that December 2014 would around the time we might hear about it. So we'll go through at least one more election in my own HD137 with a single voting precinct that contains one registered voter.


HISD 2014 Redistricting: The ‘North Forest’ Public Hearing

Reporting from Shadydale Elementary in the what used to be the heart of what used to be North Forest ISD. I can't claim to have had a perfect count of officials and former officials in the building last night, but State Rep. Senfronia Thompson and HISD Trustee Rhonda Skillern-Jones were definitely present. Attendance for the night, according to my so-so crowd-counting abilities was maybe just north of 40 people. There was a slightly bulkier handout at the meeting (scanned here). My scan doesn't pick up the maps terribly well, but those can be found in the original doc presented to the board here.

As stated yesterday, I believe this should, in theory, be an easy round of redistricting. North Forest only adds 52,368 people into Houston ISD. The average district size - post-annexation - will be 151,726. So, on the surface, there's no easy way to have a district that retains much of North Forest's region. Even if you found a way to leave NFISD whole, it would comprise just more than a third of the hypothetical district. Of course, it was also pointed out that you could triple the number of trustees on the school board. But apparently that number is set in state law, so the school board has no control over it.

As far as the timing of the map-making, the goal is to take something to the board for their August hearing. But that's not a hard timeline if there's any difficulties is an issue here. From the tone of the community feedback, it didn't seem as if there would be much of an issue. Having drawn the lines as recently as 2011, district voters have cast votes and chosen their trustees.

For the sake of discussion, however, HISD's "Priorities and Principles for Redistricting" have two criteria that add some inertia to the process:

(3) The new HISD Redistricting Plan will be based, to the extent possible, on the existing trustee district composition.

(9) Recognizing the value of incumbent-constituency relations, the new HISD Redistricting Plan will seek to keep existing trustees in their existing districts.

In short, advantage incumbents ... which seems like a big deal since North Forest won't have the possibility to contribute an incumbent to the board. I normally hate criteria like this in years that end in one. But lacking much of a dog in this fight, it would be interesting to see if 'll be interesting to see if this came into the conversation at any point. But last night? Nothing.

I was told that there might be a little rabble to be roused at this particular hearing. That explains why I just put up with rush hour traffic out of downtown to get here. I believe the anticipated feedback would along the lines of "why couldn't you leave NFISD whole?" There was a little of that, but nothing that looked particularly incendiary.

On an additional note of redistricting criteria, I'll point out another pet peeve of my own:

(6) The HISD Redistricting Plan will use whole county voting precincts, whenever possible, to draw trustee districts.

Emphasis on "whenever possible." I don't know if there is any legal basis for determining a meaningful cutoff point for that. In practice, however, it looks a little too convenient. Without a criteria for when "whenever possible" hits a brick wall, it serves to benefit the map-makers. In terms of how it played out in this process, page 27 of the new info packet highlights that fewer precincts are split in the proposed plan.

A couple of new wrinkles for this mid-decade process:

The map-makers created an "in between" map that adds NFISD to HISD and determines a new configuration. In part, this is necessary in order to assign the new turf to existing districts. Officially, those folks now have one of two new school board trustees. Also of importance is that the new map is what legally shows new population deviations between districts as well as a substantially high top-to-bottom deviation between the most-populated and least-populated district. You can't redistrict without demonstrating that there is more than 10% difference there. And the new top-to-bottom deviation is 25.2%.

Now, I don't think there is any way to add NFISD and keep those deviations within 10%. But the 25% result is, in one sense, and artificial result. There are a number of ways that NFISD could have been split up. Furthermore, there doesn't appear to be any legal basis for not redrawing lines in existing districts if it fits within the criteria put forth by HISD. Of course, you get a much bigger can of worms with the trustees, not to mention possible public pushback if you stand accused of fitting in NFISD without the traditional redistricting process. In short, though, there was a map that didn't need public input to approve and there is now a proposed map that does. For now, that just strikes me as interesting.

The numbers used for redistricting are still the 2010 Census data, but the Rice U. team working the GIS did look at ACS data to see where the districts might be headed in the future. I've got a request in for some shapefiles and I plan to do a couple of things with that. One is to overlay that with my favorite view: CVAP Majority, as well as ask the Lege Council in Austin to show the CVAP numbers for the proposed map.

At some point, I'll also look at the NFISD boundaries on top of that CVAP Majority map and possibly do the same for Total Population and Voting Age Population. That may help demonstrate some of what was mentioned at the meeting about why NFISD is split the way it is between Districts 2 and 8. In short, one part is more predominantly African-American and another isn't. Allegedly, adding all of NFISD into District 2, for instance, the AfrAm population share would drop and result in a red flag with the Justice Department. On the surface, there's a bit of irony here: you can't submit a map that retrogresses AfrAm population, especially when you're annexing a district as heavily AfrAm as NFISD. But if you were to incorporate NFISD by keeping it whole, you would retrogress the plan. This is exactly the sort of thing that passes for humor at Redistricting Meetups ... or so I hear. Part of the problem is that for all its reputation as being heavily African-American, the total population in NFISD is only 66% AfrAm and 30% Hispanic. You might be able to work it so that the AfrAm majority District 2 boosts its AfrAm numbers, but there is still a resulting impact on the Hispanic districts based on all of the other threads you end up pulling from the proverbial sweater to balance out population. Again, this is precisely the sort of thing that makes life interesting for people that play (and work) with maps.

One final not-unexpected item from the public input was a pointed question about what the Anglo share of HISD students was and what the Anglo share of HISD Trustees was. The answers are 8% and 45%. You can probably use your imagination to figure out what the intent of the question was. But it raises an interesting point in my mind about future legal challenges in redistricting. For instance, what if you compared the demographic difference between HISD-eligible parents or families and used that to challenge whether a redistricting plan was representative. I don't know that the argument would be strong enough to win a judicial or DOJ challenge. But to the extent that any jurisdiction is going to go through the redistricting process during the coming years, it will be interesting to see how differences between electorates and affected individuals plays out. In theory, that would be a very similar argument made by those who wish to see legislative redistricting done based on CVAP. Never let it be said that politics can't make for interesting bedfellows.

The next hearing is on July 8th at Pin Oak Middle School - just far enough inside the loop to not qualify as SW Houston, but this will be as close as it gets for us. Page 23 of the new packet shows the affected areas where the boundaries change. Our side of town doesn't have much. I'm expecting a civil affair. In short, if there wasn't a major hue and cry from North Forest about the draft plan, I'd be willing to bet a Coke that the proposed plan sails through in August.


HISD Redistricting – The 2014 Round

Who's ready for some more redistricting?

This caught me by surprise last night (H/T Stace). Apparently, it's due to the annexation of North Forest ISD. You can flip through the information handout here. The big picture of the changes is below. Gene Locke from Andrews & Kurth returns to be the lead legal eagle on this. But instead of working with UH and Dr. Richard Murray as the demographer, they're now using a team from Rice led by Dr. Bob Stein.

As is the norm, there will be public meetings. And the first one is up tonight:

  • Tuesday, July 1 - 6:30pm - Shadydale Elementary School - 5905 Tidwell
  • Tuesday, July 8 - 6:30pm - Pin Oak Middle School - 4601 Glenmont
  • Thursday, July 10 - 6:30pm - Hattie Mae White Educational service Center - 4400 West 18th Street
  • Tuesday, July 15 - 6:30pm - Austin High School - 1700 Dumble

I'm hoping to hit all four of these and I'll be sure to update on each. I don't see any boundary changes in my little sliver of the world, although I did get around to noticing that my recent move into the heart of Gulfton puts me in Harvin Moore's district, where I had been in Mike Lunceford's district previously. Page 18 of the info packet highlights where the geographic shifts are between districts.

At first glance, I don't notice any major demographic alterations. Two of the Hispanic-majority districts gain Hispanic population (Districts 1 and 8) while the third (District 3), drops from 77.5% VAP to 75.5% VAP. Among the African-American majority districts, there's minor wiggle room in two, while District 2 gains a VAP majority under the new map - up from 43.3% in the existing map.

All in all, fairly minor tweaks for a substantial addition in territory and population. We'll see if there's any opposition once we get into the public hearings. It might be a stretch, but typically the first target for opposition in Latino communities are districts that have high Hispanic percentages short of a majority. District 2 - an AfrAm-majority district - would start as the primary suspect, but the new map takes their VAP Hispanic numbers down from 40.3% to 36.9%. That looks like a decent move in the aggregate.

The current map:

The proposed map:

HISD Board of Trustees - June 12th presentation


The Lovers, The Dreamers … and Maps!

» Washington Post: For Maryland Democrats, redistricting referendum forces a look in the mirror
» Washington Post: Maryland ad war coming over same-sex marriage vote
» Washington Post: Costs, benefits of Md. Dream Act hard for voters to measure

I foresee a lot of interesting post-election analysis out of Maryland this season. That is all.


Redistricting Harris County State Reps: First Draft

File this under "Ain't Gonna Happen," but this is the first semi-reasonable basis of a plan that I landed on after doodling around with Harris County State Rep districts. The assumption here is that there will still be 25 seats. But the dicier assumption is that Beverly Woolley draws herself out. For that reason, I don't see this as being likely. At least not until I hear her retirement announcement.

But this draft accomplishes the tasks of protecting Legler and Bohac and a Baytown-based district. What's new about this district is a Willowbrook-area district. That's the way this map factors in the population growth on that side of town. What's curious is that if such a new district were to sprout up in that area, it would mean that another Rep would have to go. I don't see a way of mapping it out so that each of the incumbents' homes are included in some magical tortilla strip to remain safe. The likelier target for this would seem to be Legler since dropping his district makes it easier to add population to Alvarado and Hernandez. As this draft stands, several of the Dem districts on the north side of town are woefully under-populated. I think that's an inevitable bump in how the map-making proceeds in Austin. So it'll be interesting to see how the ultimate, final map deals with it all.

Politically, this is drawn to be somewhat realistic in that I don't see the GOP leadership in the county delegation drawing anything that has less than 13 GOP seats. But in a 2006/2008 style election, the map could easily become a 16-9 Dem map. What makes that possible, if the map were to sail through and become reality, is that Sarah Davis has a hard time being drawn significantly more safely than the current District 134 really is. Likewise, I think Districts 126 and 132 would look pretty interesting in a good Presidential year. 2010 is the GOP high-water mark for each district and the demographics are still changing in each of them. 132 could be drawn stronger, splitting the growing Democratic vote on the west side of town. The only thing working in favor of keeping 126 red is that Patricia Harless is on the Redistricting Committee.

But all of that raises one question in my mind: would it not be the smartest thing the GOP could do if they drew 12 pretty darned safe seats instead of shooting for 13? The demographics of Harris County are going to push the GOP out in some significant ways this decade.

Legally, there's an issue with Farrar's district. She'd have no problem being elected out of that district. But after her, the growing Anglo population in the Heights makes it less likely that Hispanic voters have a strong enough situation to elect a Representative of their choosing. It's not dissimilar to the results seen in the City Council District H special election in 2009. It's also symptomatic of the decrease in Hispanic population seen in the current iteration of Farrar's district. The challenge is this: does this mean that the Heights is going to have to be drawn in with Bohac's district? Coleman's Afr.-Am. numbers are already on the low side, so it's unlikely that you could add Montrose to his district.

There are two drivers that I see making the drawing a challenge ... the diffusion of minority voting age population places a bind on the GOP's ability to do another HD133/HD126 drawing in the present map. Those minority populations are now too valuable and necessary for VRA districts in order to keep a redistricted map VRA-compliant. But the same phenomenon also places Hispanic and African-American boundary lines at risk due to a large number of precincts being roughly 40% Hispanic and 40% African-American. In other words, it's not hard to draw an appropriate number of Hispanic districts in theory ... but the political realities that go along with redistricting are likely to make it more contentious.

Anyways, here's the draft map. Number for Voting Age Population and 2010 Governor election math is below.

Perry   White                     Population  Dev.   Hisp.   Anglo   Afram   Asian
59.7%   38.6%  126  Harless     R   166,301  -0.80%   25.9%   46.7%   18.0%   7.9%
63.9%   34.2%  127  Huberty     R   163,413  -2.52%   19.4%   64.0%   12.1%   3.0%
59.8%   38.3%  128  Smith       R   161,991  -3.37%   30.6%   55.1%   11.7%   1.3%
56.2%   41.9%  129  Davis       R   167,855   0.13%   19.7%   62.8%    6.8%   8.9%
64.9%   33.4%  130  Fletcher    R   168,764   0.67%   24.6%   54.3%   11.8%   7.9%
20.6%   78.2%  131  Allen     D     174,564   4.13%   42.7%    9.1%   40.2%   6.9%
59.5%   38.6%  132  Calegari    R   169,641   1.20%   39.6%   38.1%   14.0%   6.8%
62.3%   34.9%  133  Murphy      R   165,313  -1.39%   18.4%   58.0%   11.3%  10.4%
51.7%   46.9%  134  Davis       R   166,183  -0.87%   12.4%   71.4%    4.0%  10.6%
56.8%   41.5%  135  Elkins      R   171,150   2.10%   44.1%   36.9%    8.2%   9.5%
60.1%   38.4%  136  Bohac       R   166,362  -0.76%   35.9%   51.4%    5.7%   5.7%
39.3%   58.8%  137  Hochberg  D     173,074   3.24%   53.6%   18.7%   15.9%  10.3%
63.5%   34.7%  138  OPEN        R   164,240  -2.03%   23.6%   52.9%   11.8%  10.1%
19.6%   79.3%  139  Turner    D     147,490 -12.02%   39.1%   12.4%   42.9%   4.7%
37.5%   60.9%  140  Walle     D     157,662  -5.95%   79.2%   11.6%    7.8%   0.8%
37.9%   60.7%  141  Thompson  D     151,869  -9.41%   38.3%   12.1%   44.7%   3.7%
20.0%   79.1%  142  Dutton    D     145,031 -13.49%   30.8%   14.1%   52.9%   1.2%
42.1%   56.5%  143  Hernandez D     163,235  -2.63%   61.6%   20.1%   15.7%   1.7%
52.8%   45.3%  144  Legler      R   165,181  -1.47%   58.3%   34.4%    4.3%   2.1%
29.6%   68.7%  145  Alvarado  D     170,005   1.41%   85.9%    9.4%    2.8%   1.3%
24.7%   74.0%  146  Miles     D     168,300   0.40%   20.6%   23.8%   45.9%   8.3%
26.5%   72.1%  147  Coleman   D     164,426  -1.92%   34.7%   19.5%   37.0%   7.5%
34.2%   63.7%  148  Farrar    D     146,877 -12.38%   38.8%   45.4%   11.0%   3.2%
42.5%   56.1%  149  Vo        D     170,821   1.90%   35.6%   15.1%   25.7%  22.0%
68.4%   29.7%  150  Riddle      R   162,577  -3.02%   17.5%   67.3%    7.5%   6.1%

Second Effort at a Full Citywide Redistricting Plan

Just in case anyone stumbling onto the blog is looking for "those maps," here's a revised version of the original I posted. It's a little bit cleaner in terms of neighborhood contiguity, but still has some potential questions about political feasibility.

A bit of a repeat point from the original map, but my bias are as follows:

- Find a way to create a third Hispanic seat. This revision has Ed Gonzalez in the northern (red/orange) district and I think he'd win it despite the loss in Hispanic voting strength.

- Maximize the Asian population within a new District F.

- Create an additional minority district as one of the new districts. Fort Bend County strikes me as a worthwhile candidate for this, but I'm not 100% married to it being centered there.

- Montrose + District C ... and Heights, if possible.

The feasibility of that third district is going to be interesting to watch as the process goes forward. Both the southern and northern Hispanic districts would clock in at under 50% Hispanic in terms of CVAP population and likely electorate in a city election year (especially a super-low turnout year like this year should be). I may be on the optimistic side, but I don't see either Gonzalez or Rodriguez having much problem for the duration of their council tenure. The "new" East End district is the most secure opportunity for Hispanics to elect a candidate of their choosing. Part of the argument I'd put forward for the incumbent-occupied districts is that they each include enough non-Anglo voters to be able to have a majority non-Anglo CVAP. Also, the Hispanic numbers would be likely to grow over the decade. Obviously, I'll be wracking my brain with this configuration once the City opens up their redistricting kiosk. But, generally, I think there's something to like about seeing a map that gives three very distinct Hispanic communities a strong voice in city elections.

It should be evident that District E has a bit more taken out along the eastern border as I believe it may be needed for contiguity. I'm still hoping there would be a way to put some of that back into District I in order to boost the Hispanic numbers. Another negative side-effect of that move is that Dist. I eats into District D a bit now - that's something that I think will be argued against as many of those same folks aren't happy with being in either I or E. Spring Branch is a little bit more whole in the new A. And yes, the new C contains Ellen Cohen's home precinct.

Population-wise, this checks out against the 2008 estimates. I'd post demographics, but I'll probably have a plan developed on the city's setup within a week or so. So I'll save the full details until then, if that's cool.

Questions? ... complaints? That's what comments are for.

Zoomed In:

Zoomed Out:


Houston Challenges 2.099M

» Chron: City to challenge census count

Mayor Parker throws a yellow flag onto the field, essentially asking for a recount ...

The city of Houston will ask the U.S. Census Bureau to change its official count, raising questions about whether some apartment complexes or even entire neighborhoods were missed.

Houston's population is 2,099,451, according to Census data released last week. That's more than 100,000 fewer people than earlier estimates, and slightly below the 2.1 million that triggers an expansion of City Council to 16 members.

The expansion is still on, as city planners and independent researchers try to determine what went wrong.

I think it's probably a no-win situation that the City has to put up with here. If you don't expand, you get sued by one side. If you do expand, you get sued by the other. For my view, it's better to plan forward for the decade and add the districts now. If you hold off until 2 more years, the end result would simply be that you divide districts by 11 with the same data everyone is looking at now.


DIY Redistricting

» The Tennessean: Do-it-yourself redistricting gives citizens a voice

Have I mentioned what a great tool Dave's Redistricting App is? Well now the Tennessean is in on the secret.

On his own time, Bradlee built "Dave's Redistricting App," an online mapping tool. It's free, and anyone can use it. You choose a state, decide how many districts to slice it into and then click away, coloring the map into lots of tiny pieces. As you draw your own congressional or state legislative districts, the app spits out Census data on each one's population and racial composition. With a little persistence, anyone can produce his own redistricting plan.

Bradlee quickly discovered he wasn't alone in his passions. Since the app went live in 2009, hundreds of people have used it to draw political maps; the site is now live for every state but Alaska.

You can check the "Recent Diaries" sidebar on Swing State Project to see a constantly-growing list of user-generated maps and redistricting plans. I can't state that it's the most productive hobby in the world to generate maps like that, but it's incredibly instructive in learning about a state or region. Most of those are created for Congressional maps, but you can create maps for any single member districts if you have a ballpark idea of what to have for population per district. My own backgrounder on using the tool was posted here after a little bit of interest was generated at a redistricting forum put on at St. Thomas University.

For a newsy/opinionated take on the open-ness of redistricting, I'll suggest this read by Michael McDonald and Micah Altman from last July. Among the points I think they raise are the fact that states and localities have been slow to open up the software needed to allow citizens to redistrict on their own. Understandably, cost has been a factor as budgets are beyond tight these days. The City of Houston will allow anyone who can make it to the Planning Department the opportunity to create their own district and submit it for consideration. Texas allows the same if you can make it to Austin. But while the software for doing so is made freely available to legislators on their state-issued laptops, it is not offered for download by private citizens.

I am working on a project to allow for greater ease in exploring census data in Texas. It's not quite the same as a redistricting application and I doubt that I'll get around to coding for that. But if it's of any use or interest to anyone, is a work in progress that might be of interest. I've found it helpful in answering a number of questions that go into redistricting and it's helpful in a number of other areas, as well. If you've got any feedback on making it better and more user-friendly, feel free to drop a comment or an email my way. Consider yourselves beta-testers. I'll be updating the 2010 Census data as soon as I've got it databased.


Let the Lawsuits Begin!

» Texas Tribune: Texas Redistricting Lawsuit: Count Citizens Only

Seems early for this, but I guess there's no time like the present:

Attorney Michael Hull of Austin, representing three North Texas voters, sued the state and a bunch of others, alleging that counting undocumented immigrants in political districts has an unfair and illegal effect on voters in districts with smaller numbers of non-citizens.

The logic goes this way: If two districts have the same populations and one has more non-citizens than the other, it takes fewer voters in that district to swing an election. Fewer citizens means fewer voters means a smaller number makes a majority. Each vote is, compared to the district with more citizens, worth more.

There's a variety of ways that I think that either the lege or rightwing legal activists will test redistricting this year and this was one of the expected routes. As protective as GOP electeds have been of the judiciary during years of Democratic administrations, I wouldn't be surprised to see a 3-judge panel move this along. Where it ends up, though, may be more of an open question. Either you accept that representation is determined by population, or you start whittling away at what the Constitution considers an American population.

So what's left to run through the courts in terms of abstract arguments like this? I haven't dived deep into the legal side of redistricting yet, but I don't seem to recall there being something that really challenges the Voting Rights Act - namely Section 5. That should be expected somewhere, if not Texas. But the more impactful challenge will be to see which entities attempt a judicial bail-out of preclearance. That one, I'd expect the lege to try with our maps here in Texas. Whether that is accompanied by any significant regression in minority representation remains to be seen.

Fatal error: Call to undefined function xwp_footerx() in /nfs/c01/h02/mnt/4748/domains/ on line 18