The 248 complaints of 2014 that describe eight years of dashboards

The date the paperwork arrived The self-selected sample The count that measures the looking

Dark navy slide headed “Field quality review · monthly pack · demonstration slide” on the left and “2007 model year · 459 consumer complaints · NHTSA component: structure” on the right. The headline reads “The dashboards started failing in 2014.” A bordered gold chip beneath it reads “Demonstration · drawn by us from the federal complaint database”. The standfirst says every complaint the regulator holds for this model year under this component code is on the slide — 459 of them, none dropped, one make, one model, one model year, one component, no sampling and no smoothing; that the count axis is linear and starts at zero with nothing truncated, broken, clipped or logged; and that eight quiet years were followed by 248 complaints in 2014 alone, 6.9 times the year before and 54% of everything ever filed under this code. The main panel, “Complaints about the 2007 dashboard, by year”, subtitled “Count of complaints on the federal register · component ‘structure’ · linear count axis starting at zero · every complaint held for this model year is in the chart, none dropped, nothing smoothed or rescaled”, is a column chart on a count axis running 0 to 250 with gridlines every 50. Twenty-one columns run from 2006 to 2026 reading 3, 6, 8, 4, 15, 6, 15, 36, 248, 35, 6, 18, 16, 15, 9, 5, 6, 2, 4, 1 and 1. All are blue except the 2014 column, which is gold and towers over everything else, labelled 248, with an annotation beside it on two lines reading “248 complaints in one year” and “against 36 the year before”; the final 2026 column is hatched to mark a part-year. Every other column is a stub along the bottom. Three panels sit below. “Inside the spike”, subtitled “Complaints per month, 2014”, is twelve columns reading 3, 5, 5, 7, 10, 18, 29, 66, 41, 36, 17 and 11, with August picked out in gold at 66. “Share arriving in 2014”, subtitled “Percent of all 459, zero-based axis”, is two horizontal bars — 2014 at 54.0% in gold and all other years at 46.0% in blue — under a line reading “One year holds more than half the register”. “The review, tabulated”, subtitled “Counts read straight off the register”, is a monospace list giving 2012 complaints 15, 2013 complaints 36, 2014 complaints 248, year on year ×6.9, peak month Aug · 66, then a rule and all years all complaints 459, share filed in 2014 54.0%, complaints dropped 0, status ESCALATE, closing with a note that the count rose almost sevenfold in a year and fell back just as fast and is treated as a 2014 build-quality event and escalated to the supplier review. Five cards run down the right. “The finding, in one figure” prints a large gold ×6.9 beside “complaints in 2014 against 2013 — 248 against 36, the sharpest rise anywhere in the series”. “Read-out for the review” says a step change this large in a single year is a manufacturing event rather than drift and recommends pulling 2014 build records for the dashboard line and opening a supplier review. “The set as read” tabulates make and model Toyota Camry, model year 2007, component STRUCTURE, complaints 459, span of register 2006–2026 with an asterisk, mentioning “dash” 377, mentioning “dashboard” 323, sampling none. “Why nothing here is bent” notes one population, one unit, one component, a linear count axis from zero, every year drawn at the same width, every complaint in the chart and the 2026 column marked as a part-year, ending “Whatever is wrong with this slide, it is not the drawing.” “How the years were counted” explains that each complaint is placed in the year its record was received by the regulator — the date-filed field, grouped by calendar year — and that every chart in the pack is built the same way. Footer chips read “Every complaint included”, “One model year, one component”, “Linear axis from zero”, “Nothing smoothed” and “Nothing dropped”, above a footnote naming the source, giving the counts of records containing “dash” and “dashboard”, listing the yearly and monthly counts, noting that the register was pulled on 25 September 2026 so the 2026 column is a hatched part-year, and stating that the review, the status line and the recommendation are invented while every figure behind them is the regulator’s own. — Showing the misleading chart

The same 459 complaints on a cream background, kicked “The same 459 complaints · read by the date the dashboard failed” and “2007 model year · NHTSA consumer complaints · component: structure”, headlined “The 248 complaints of 2014 that describe eight years of dashboards.” A badge reads “Demonstration · both charts drawn by us from the federal complaint database”. The standfirst grants that the count is right and the axis linear and zero-based, and says the x-axis is the problem: it is the date the owner filled in the form rather than the date the dashboard failed, and the register carries both dates on every record. Read by the owner’s own date the cliff becomes a slope — 15, 18, 16, 34, 31, 59, 83, 106 across 2007 to 2014 — and of the 248 filed in 2014, 153, or 61.7%, describe a dashboard that had already gone sticky in 2013 or earlier, one of them as far back as 2006. The main panel, “One set of complaints, two dates”, subtitled “The same records, counted by the year the complaint was filed and by the year the owner says the dashboard failed · linear count axis starting at zero · nothing added; one record whose incident date is 1996 is named in the footnote instead of plotted, and 2026 is a part-year”, plots two lines from 2006 to 2026 on a count axis running 0 to 250. A dark red line, “By year filed — the first chart’s axis”, sits along the bottom and then spears up to 248 at 2014 before collapsing back. A blue line, “By year the owner says the dashboard failed”, climbs gradually through 6, 15, 18, 16, 34, 31, 59, 83 to a broad peak of 106 at 2014 and then falls away. A dashed tan rule joins the two 2014 points, annotated “248 filed in 2014” above “the first chart’s spike” in red at the top, and “106 dashboards dated 2014” above “by the owner’s own date — a slope, not a cliff” in blue below. Three panels sit underneath. “Where the 2014 filings came from”, subtitled “The 248 complaints filed in 2014, by the year the owner says the dashboard failed”, is nine columns reading 1, 4, 9, 9, 18, 20, 32, 60 and 95 for 2006 through 2014, the first eight in blue and 2014 in khaki, headed “153 of the 248 — 61.7% — predate 2014”. “How long owners waited”, subtitled “Filed minus failed, 458 records · zero-based”, is four horizontal bars — under 90 days 36.7%, 90 days to a year 16.8%, one to three years 29.5%, over three years 17.0% — under a line reading “Median 288.5 days · 45.6% waited over a year”. “Read-out”, subtitled “The same register, the other date”, gives 2014 by filing date 248 complaints, ×6.9 on the year before; 2014 by failure date 106 complaints, ×1.28 on the year before; median wait 288.5 days between the fault and the form; and 45.6% of complaints filed more than a year after the date the owner gives — closing that the spike is not a year in which dashboards failed but a year in which owners wrote in, that the timing is coincidence rather than proof, and that it sits in the months the sticky dashboard was on Florida television and months before any warranty letter was posted. Eight cards fill the right in two columns. “A complaint is a volunteer” notes that nobody sampled these owners — each found the regulator’s website, typed in a VIN and wrote a paragraph — so the count measures how many people were moved to do that, and what moves them is mostly whether they had just heard the problem had a name. “What happened in the summer of 2014” records filings of 29 in July, 66 in August, 41 in September and 36 in October before a return to the old baseline, and investigations into melting dashboards that summer by Florida stations including WPTV in West Palm Beach and NBC 6 in Miami, with the cars seven years old and unchanged. “The warranty programme came afterwards” notes that Toyota’s ZE6 warranty enhancement, covering cracked or sticky dashboards on the 2007–2011 Camry among five model lines, is set out in a warranty policy bulletin dated 10 December 2014 recording the reimbursement phase as announced that mid-December, after the spike had peaked and fallen. “Re-bucketing does not cure it” warns that the blue line is better rather than clean, since it still contains only dashboards whose owners eventually wrote in, so its own fall after 2015 is the wave of attention ending rather than dashboards recovering. “The denominator nobody has” asks 459 complaints out of how many 2007 Camrys still on the road, notes the register does not say and the fleet shrinks every year, and concludes that a count with a falling, unmeasured denominator cannot carry a rate. “What to do instead” says to plot the date the event happened rather than the date the paperwork arrived and to say which was used, to put the known attention events on the chart as rules, and then to give the count a denominator or call it a count. “Where else this hides” lists adverse-event reports after a drug is in the news, bug trackers after a blog post, safety near-misses after a safety campaign and harassment reports after a policy launch, noting that the encouraging thing to do puts a spike in the series. “How this one was drawn” records that both charts come from one download of the federal register and that the only difference between them is which date field goes on the x-axis. Footer chips read “The same 459 complaints”, “The same register”, “Linear axis from zero”, “Only the date field changed” and “One record set aside, and named”, above a footnote giving the retrieval, both full series, the one record set aside and why, the incident years behind the 2014 filings, the waiting-time bands and the ZE6 bulletin reference. — Showing the misleading chart

Be the first to star this exhibit

A quality pack we drew counts every complaint the US vehicle regulator holds against one model year under a single component code — 459 of them, 323 of which mention a dashboard, none dropped, on a linear count axis from zero — and finds eight quiet years followed by 248 complaints in 2014, almost seven times the year before. Nothing in the drawing is bent. The x-axis is the date the owner filled in the form, and the register carries a second date on every record: the day the owner says the dashboard actually went sticky. Count by that one instead and the cliff becomes a slope, because 153 of the 2014 filings — 61.7% of them — describe a dashboard that had already failed in 2013 or earlier, one of them in 2006. The spike sits in the months the problem was in the news rather than in the year the dashboards failed: the timing is coincidence rather than proof, but the monthly shape is the shape of attention and not of a production line.

US National Highway Traffic Safety Administration, Office of Defects Investigation — consumer complaint database, retrieved 25 September 2026 via api.nhtsa.gov (make Toyota, model Camry, model year 2007, component “STRUCTURE”). Warranty programme dates from Toyota warranty policy bulletin POL14-11, 10 December 2014, hosted by NHTSA. The 2014 press coverage is reported by Florida stations WPTV, West Palm Beach, and NBC 6, Miami. Both drawings are ours., 2026Added September 25, 2026

01The claim

The dashboards started failing in 2014. Every complaint the regulator holds for this model year and this component is on the slide — 459 of them, one make, one model, one model year, one component, no sampling, no smoothing and none dropped. The count axis is linear and starts at zero; nothing is truncated, broken, clipped or logged, every year is drawn at the same width, the 2026 column covers a part-year and is marked as one, and each complaint is placed in the year its record reached the regulator, the same way every chart in the pack is built. Eight quiet years — 3, 6, 8, 4, 15, 6, 15, then 36 — and then 248 complaints in 2014 alone. That is 6.9 times the year before and 54% of everything ever filed under this code, arriving inside a single twelve-month window, with 66 of them in August alone. Read-out for the review: a step change this large in one year is a manufacturing event, not drift. Recommendation: pull the 2014 build records for the dashboard line and open a supplier review. Status: escalate.

02The trick

The drawing is doing nothing. One population, one unit, one component; a linear count axis from zero; every year the same width; every complaint the register holds in the chart. The slide survives any audit of its geometry, because what is wrong with it is not in the picture — it is in the x-axis, which is not the axis the headline needs. The field behind it is the date the complaint was typed into the regulator’s form, and the register carries a second date on every single record: the day the owner says the dashboard actually went sticky. Both were available; the chart uses the one that has nothing to do with the car. Plot the same 459 complaints by the owner’s own date and the cliff becomes a slope — 6, 15, 18, 16, 34, 31, 59, 83, 106 from 2006 to 2014, a 1.28-fold rise into 2014 rather than a 6.9-fold one. The arithmetic behind the difference is on the records themselves: of the 248 complaints filed in 2014, 153 of them — 61.7% — describe a fault that had already appeared in 2013 or earlier, stretching back to one dated 2006, and only 95 describe one that appeared in 2014. (236 of those 248 mention a dashboard; the component code they all share, “structure”, is a little wider than the story, which is why the counts here are of the code and not of the word.) Across the set the median gap between the fault and the form is 288.5 days, and 45.6% of these owners waited more than a year. Which leaves the question of what 2014 actually was, and the monthly shape answers it: 29 filings in July, 66 in August, 41 in September, 36 in October, then back to the old baseline. That is not how a production line fails; it is how a news cycle runs. Florida stations — WPTV in West Palm Beach and NBC 6 in Miami among them — spent that summer investigating melting dashboards, and Toyota’s ZE6 warranty programme, which covers cracked and sticky dashboards on the 2007–2011 Camry among five model lines, is set out in a bulletin dated 10 December 2014 recording the reimbursement phase as announced that mid-December, after the spike had already peaked and fallen. None of that is proof: the register cannot say why any one person filed on any one day, and the coverage and the complaints ran together rather than in a provable order. What it does establish is that the programme followed the spike rather than causing it, and that the months the filings arrived in were months the problem was being talked about. Underneath all of it sits the thing that makes a complaint register different from a survey: nobody sampled these owners. Each one had to find the website, enter a VIN and write a paragraph, so the count measures how many people were moved to do that — and what moves them is mostly whether they had just heard that their private annoyance had a name, a cause and possibly a remedy. (The slide is our demonstration in the manner of an internal quality pack: the review, the status line and the recommendation are invented, and every figure behind them is the regulator’s own. Nothing here turns on whether the dashboards were defective. Toyota later extended warranty coverage for the condition; whether it should have, and when, is not what this slide is about. The argument is with the axis, not with the cars.)

03The fix

Put the date the event happened on the axis, not the date the paperwork arrived, and say in the caption which one you used — because in any register built from reports there are always at least two dates, and the one the database sorts by is almost never the one the question is about. Here that single substitution turns a sevenfold step into a gentle climb, and it costs nothing: the records already carry the field. Then draw what you know about the looking. Mark the broadcast, the filing, the recall letter, the policy launch as rules across the plot, so a reader can see which part of the series is the phenomenon and which part is the publicity; a spike that lines up with an attention event to the month has explained itself. Be honest that the better axis is not a clean one. The redrawn line still contains only the dashboards whose owners eventually wrote in, so its own decline after 2015 is the wave of attention ending rather than dashboards recovering — re-bucketing exposes the artefact and does not remove it, and saying so is the difference between a fixed chart and a chart that has merely moved its problem somewhere less visible. Where you can, anchor the volunteers against something drawn: a random survey of owners, a sample of vehicles inspected on a schedule, warranty claims from the dealer network, anything at all that did not require the owner to act. A few hundred cars inspected on a schedule will settle in an afternoon what these 459 volunteered reports cannot settle at all. Give the count a denominator or call it a count: 459 complaints out of how many of these cars are still on the road is a rate, and the register does not know the answer, so a chart of raw counts against a fleet that shrinks every year is answering a question nobody asked. And title the chart with the thing it counts: “complaints received” instead of “failures” would have made this whole slide impossible to misread.