Is Poland
really developing?
439 strategic investments, 266 of them current, with 173 earlier landmarks added as background. Each is worth at least PLN 500 million or has been deemed strategic. Completed ones, ones under construction and the ones that failed, from governments of every stripe. Zero politics, sources only. Judge for yourself.
01Now
Right now 102 major investments are under way.
Roads, rail, factories, ports and power plants are being built now, all at once and across the country.
See all 102 in the filterable list
Records in progress that are at the construction or commissioning stage. Announcements, permits and tenders are left out, however close their stated deadline. The order follows momentum, which combines the build stage with proximity to opening.
02Leap
What we’ve already built.
What already stands and works comes to 76 completed megaprojects of the contemporary era, worth about PLN 135 billion, across a dozen fields at once. Motorways and express roads grew 7.6× in one generation, from nearly 720 km in 2004 to almost 5,466 km in 2025.
show all sectors (+12)
The bars show the value of completed contemporary-era projects by sector. The catalogue holds 238 records with status “completed”. Earlier landmarks are collected selectively, so this chart leaves them out. The line between the eras is an editorial one, and the delivery date does not decide it. Figures are nominal, in prices from various years and not adjusted for inflation. Amounts recorded in a foreign currency are converted at the fixed rate of EUR 4.30, USD 4.00 and SEK 0.38. Nuclear and offshore wind are left out and arrive here only once delivered. The sum is “value-by-stage”, a different measure from “total cost”. Express roads: nearly 720 km in 2004 (GDDKiA, 2004 figures), almost 5,466 km in 2025 (GDDKiA, 2025 figures).
Where these figures come from
The PLN 135 billion combines figures on different bases. PLN 91.6 bn is in announced budget envelopes (68%, 46 rec.). PLN 24.7 bn is in contract values (18%, 16 rec.). PLN 19.0 bn is in estimates (14%, 10 rec.). These amounts are not settled outlays. Every card states its own basis for the figure.
show the source quotes
„Wchodząc do Unii Europejskiej, 1 maja 2004 r., Polska miała blisko 720 km dróg szybkiego ruchu”
GDDKiA, 2004 · GDDKiA – Wspólnota europejska i polskie drogi
„Obecnie kierowcy mogą korzystać z niemal 5466 km dróg szybkiego ruchu, w tym ponad 1894 km autostrad i 3571 km dróg ekspresowych.”
GDDKiA, 2025 · GDDKiA – Podsumowanie 2025 roku
Why separately?
Only the current era is collected systematically. The earlier landmarks are a selection, hence their own section. Their costs do enter the “value-by-stage” figure, so the headline sum covers both eras.03What’s coming
More is scheduled for 2026–2032 than opened in 2020–2025.
The lead of 2026–2032 over 2020–2025 holds for those two windows only. Counting every opening on the axis, more are already behind us than ahead. Projects in progress or planned with no stated opening year get no bar, so the future years sit lower than the data would put them. By value, the two tallest bars fall in 2035 and 2036. Two records with declared dates fill most of them, the “Y” high-speed railway and the first nuclear power plant.
Show the figures (table)
| Year | Delivered % GDP | Planned % GDP | % of GDP | Projects |
|---|---|---|---|---|
| <1995 | 0.38% | 0% | 0.38% | 4 |
| 1995 | 0% | 0% | 0% | 0 |
| 1996 | 1.07% | 0% | 1.07% | 1 |
| 1997 | 0% | 0% | 0% | 3 |
| 1998 | 0.22% | 0% | 0.22% | 1 |
| 1999 | 0.23% | 0% | 0.23% | 2 |
| 2000 | 0.17% | 0% | 0.17% | 1 |
| 2001 | 0% | 0% | 0% | 0 |
| 2002 | 0.26% | 0% | 0.26% | 1 |
| 2003 | 0.47% | 0% | 0.47% | 2 |
| 2004 | 0.06% | 0% | 0.06% | 1 |
| 2005 | 0.2% | 0% | 0.2% | 1 |
| 2006 | 0% | 0% | 0% | 0 |
| 2007 | 0.66% | 0% | 0.66% | 7 |
| 2008 | 0.33% | 0% | 0.33% | 5 |
| 2009 | 0.4% | 0% | 0.4% | 5 |
| 2010 | 0.41% | 0% | 0.41% | 7 |
| 2011 | 1.72% | 0% | 1.72% | 10 |
| 2012 | 1.07% | 0% | 1.07% | 14 |
| 2013 | 1.26% | 0% | 1.26% | 8 |
| 2014 | 1.51% | 0% | 1.51% | 12 |
| 2015 | 1.55% | 0% | 1.55% | 22 |
| 2016 | 0.68% | 0% | 0.68% | 11 |
| 2017 | 1.06% | 0% | 1.06% | 12 |
| 2018 | 0.53% | 0% | 0.53% | 6 |
| 2019 | 1.24% | 0% | 1.24% | 9 |
| 2020 | 0.58% | 0% | 0.58% | 5 |
| 2021 | 1.02% | 0% | 1.02% | 7 |
| 2022 | 0.77% | 0% | 0.77% | 11 |
| 2023 | 1.17% | 0% | 1.17% | 21 |
| 2024 | 0.48% | 0% | 0.48% | 11 |
| 2025 | 0.98% | 0% | 0.98% | 13 |
| 2026 | 0.12% | 2.18% | 2.29% | 28 |
| 2027 | 0% | 2.01% | 2.01% | 23 |
| 2028 | 0% | 1.31% | 1.31% | 24 |
| 2029 | 0% | 0.7% | 0.7% | 22 |
| 2030 | 0% | 2.13% | 2.13% | 16 |
| 2031 | 0% | 1.69% | 1.69% | 5 |
| 2032 | 0% | 1.54% | 1.54% | 9 |
| 2033 | 0% | 0% | 0% | 2 |
| 2034 | 0% | 0% | 0% | 0 |
| 2035 | 0% | 3.18% | 3.18% | 4 |
| 2036 | 0% | 5.26% | 5.26% | 1 |
| 2037+ | 0% | 0% | 0% | 3 |
The denominator is GDP in the year a project's amount comes from, whatever its opening year. Read on
Projects opened in 2020–2025 add up to about 5% of GDP, on average 0.83% a year. The 2026–2032 window holds about 11.67%, on average 1.67% a year, and almost all of it is a declared pipeline, not yet delivered. None of these amounts is annual spending. The chart leaves out records with no stated opening year, and far more of those are under construction or planned than delivered, so the pipeline comes out low here. Future dates are declared dates, and some will slip.
How this chart is computed
The bars run by opening year. Completed projects sit in their delivery year, the rest at their current planned date. Bar height is the share of GDP, meaning cost divided by nominal GDP in the year the cost figure comes from. That lets projects from different eras be compared despite inflation and the 1995 redenomination. The denominator is GDP in the price year where the source gives one. For amounts converted from a foreign currency we take the year of the exchange rate, and otherwise the year of the cost figure’s date. Where that is missing too, the year comes from the record’s timeline, in the order delivery, planned date, start, announcement. 0 projects have a stated price year and 282 use one of the other three rules. The bar counts rest on 350 records and the height on 282, because 68 projects carry no recorded cost.
macro backdrop: data on the economy, separate from project costs
Poland’s real GDP at constant prices grew 237% since 1990, the most in the region. Czechia’s GDP grew 81% over the same period and Germany’s 55%. This figure is a measure of the scale of the economy. It measures neither prosperity nor project costs. A separate measure, GDP per capita in PPS, rose between 2004 and 2024 from 51 to 79% of the EU average. World Bank, Eurostat. GDP and PPS serve only as a backdrop here. The GDP growth figure we compute ourselves from a World Bank series. The PPS index is read from a Eurostat table, and the road kilometres carry a quote from a primary source. Only that third figure meets the bar every project field has to clear.
04Honestly
Not everything worked. And it’s good that you can see it.
Among the 92 current records that were completed or stopped, 17 did not work out, about 18%. Stopped here means cancelled, superseded or paused. A project under construction or planned has no outcome yet either way, so it is not in the denominator. A project counts as failed when it was cancelled, superseded or paused, or carries a shortfall material to the project's scale. The shortfall must be established, named by an audit or under investigation. The full list is longer (24), because it also holds failures among the earlier landmarks (7). The rule does not count delays or overruns, although there are far more of those. Where a source speaks about the schedule, 86 of 99 records ran late. Where it speaks about the budget, 26 of 31 are over it. Those two fractions run over the whole database, including records in progress and landmarks. None of these figures is a rate for Polish investment as a whole. The database is a selection, so it is no random sample. On top of that, a withdrawn announcement gets its outcome sooner than a finished build, and sources report a slip more often than a deadline kept. The “Failed” pill in the catalogue follows the same rule. Each of these figures can be reproduced from /api/projekty.json and /api/statystyki.json. We show failures with the same source rigour as successes. They are rarely black-and-white.
This is why the list reads as one-sided in time. A cancellation, a replacement or an established shortfall takes years by its nature, so any large investment active today may have no verdict yet, regardless of when it started. The Central Transport Hub (CPK) and the second nuclear plant are still under way, and the rule above cannot rule on them yet. That does not mean they are faring better. It only means nothing is settled yet.
Warsaw–Radom Airport
95,646 passengers in 2025 against 3,000,000 capacity – 3%
Ostrołęka C Power Plant
~PLN 1.3 billion on a coal unit that was never finished – construction was halted
Izera / the Polish EV
Brand cancelled (PLN 580 million); the company is preparing a different car with Foxconn under a new brand, production 2029
Cross-city rail tunnel in Łódź
The boring shield has stood still since November 2024; the cost grew from about PLN 1.3 billion net to about 2.3 billion
Orlen Olefins III (original scope)
Budget ballooned from PLN 8.3 to as much as 51 billion including infrastructure; PLN 16.8 billion write-off; superseded by the narrower “New Chemistry” (PLN 35.8 billion)
And even these are rarely black-and-white: Izera was cancelled, and the company that developed it is preparing a different car with Foxconn under a new brand. The post-Northvolt battery factory in Gdańsk is not on the list, although its investor went bankrupt: a new owner restarted production there. Every case is in the database with full context.
The 5 examples above were picked by hand, each for a different kind of failure, and the selection is no ranking of losses. All 24 records sit on a page of their own: the full “Failed” list.
05Where
Where it’s being built.
The map groups projects by voivodeship. Circle size follows the number of projects in the region, and their value plays no part in it. The point map and filters are in the “All projects” section.
Tap a region (or hover) to see its breakdown.
Circle colour is the field with the most projects in the voivodeship:Transporttie (several fields with the same count)Energy
409 projects on the map; circle size = number of projects in the voivodeship (not value). Regional counts add up to more than 409 because 65 projects span several voivodeships. The same applies to the amounts: a multi-region project enters each of them at full value, so the regional figures must not be added up – they come to about 44% more than those records do. 25 sensitive sites are counted at voivodeship level (no precise point). 30 projects have no assigned region – off the map.
06All projects
Now it’s your turn.
Start from a field
9 indexes, each with a summary and the full project list. Filters and the map are below.
439 projects, PLN 1,424 billion
The buckets divide the database with nothing left over, since 102 + 238 + 67 + 9 + 23 = 439. The last term is the stopped records, meaning cancelled, superseded or paused. The “Failed” pill (24) cuts across the buckets. It holds every stopped record and also the live or completed ones where a source states a material shortfall. Added to the other pills, it would count some records twice.
what is momentum?
Momentum = how close a project is to opening: build stage × proximity to its date. “Now” and “home stretch” rank above early plans. Two buckets sit outside that scale because they describe a state rather than a distance from opening: “partly in service” is a project whose first part already runs while the rest is built, and “failed” cuts across the other buckets instead of being one more.