diff --git a/.circleci/config.yml b/.circleci/config.yml
index 977a3517e9749..d82e97b61cb5a 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -3,12 +3,12 @@ version: 2.1
 jobs:
   doc-min-dependencies:
     docker:
-      - image: circleci/python:3.7.7-buster
+      - image: cimg/python:3.8.12
     environment:
       - OMP_NUM_THREADS: 2
       - MKL_NUM_THREADS: 2
       - CONDA_ENV_NAME: testenv
-      - PYTHON_VERSION: 3.7
+      - PYTHON_VERSION: 3.8
       - NUMPY_VERSION: 'min'
       - SCIPY_VERSION: 'min'
       - MATPLOTLIB_VERSION: 'min'
@@ -48,7 +48,7 @@ jobs:
 
   doc:
     docker:
-      - image: circleci/python:3.7.7-buster
+      - image: cimg/python:3.8.12
     environment:
       - OMP_NUM_THREADS: 2
       - MKL_NUM_THREADS: 2
@@ -100,13 +100,13 @@ jobs:
 
   lint:
     docker:
-      - image: circleci/python:3.7
+      - image: cimg/python:3.8.12
     steps:
       - checkout
       - run: ./build_tools/circle/checkout_merge_commit.sh
       - run:
           name: dependencies
-          command: sudo pip install flake8
+          command: pip install flake8
       - run:
           name: linting
           command: ./build_tools/circle/linting.sh
@@ -142,7 +142,7 @@ jobs:
             - ~/scikit_learn_data
   deploy:
     docker:
-      - image: circleci/python:3.7
+      - image: cimg/python:3.8.12
     steps:
       - checkout
       - run: ./build_tools/circle/checkout_merge_commit.sh
diff --git a/.github/workflows/wheels.yml b/.github/workflows/wheels.yml
index baffd7aedde4a..e40f3862b7d6e 100644
--- a/.github/workflows/wheels.yml
+++ b/.github/workflows/wheels.yml
@@ -36,7 +36,7 @@ jobs:
         name: Check build trigger
         run: bash build_tools/github/check_build_trigger.sh
 
-  # Build the wheels for Linux, Windows and macOS for Python 3.7 and newer
+  # Build the wheels for Linux, Windows and macOS for Python 3.8 and newer
   build_wheels:
     name: Build wheel for cp${{ matrix.python }}-${{ matrix.platform_id }}-${{ matrix.manylinux_image }}
     runs-on: ${{ matrix.os }}
@@ -51,10 +51,6 @@ jobs:
           # Window 64 bit
           # Note: windows-2019 is needed for older Python versions:
           # https://github.com/scikit-learn/scikit-learn/issues/22530
-          - os: windows-2019
-            python: 37
-            bitness: 64
-            platform_id: win_amd64
           - os: windows-2019
             python: 38
             bitness: 64
@@ -69,10 +65,6 @@ jobs:
             platform_id: win_amd64
 
           # Window 32 bit
-          - os: windows-latest
-            python: 37
-            bitness: 32
-            platform_id: win32
           - os: windows-latest
             python: 38
             bitness: 32
@@ -83,11 +75,6 @@ jobs:
             platform_id: win32
 
           # Linux 64 bit manylinux2014
-          - os: ubuntu-latest
-            python: 37
-            bitness: 64
-            platform_id: manylinux_x86_64
-            manylinux_image: manylinux2014
           - os: ubuntu-latest
             python: 38
             bitness: 64
@@ -100,11 +87,6 @@ jobs:
             manylinux_image: manylinux2014
 
           # Linux 64 bit manylinux2010
-          - os: ubuntu-latest
-            python: 37
-            bitness: 64
-            platform_id: manylinux_x86_64
-            manylinux_image: manylinux2010
           - os: ubuntu-latest
             python: 38
             bitness: 64
@@ -124,10 +106,6 @@ jobs:
             manylinux_image: manylinux2014
 
           # MacOS x86_64
-          - os: macos-latest
-            bitness: 64
-            python: 37
-            platform_id: macosx_x86_64
           - os: macos-latest
             bitness: 64
             python: 38
diff --git a/.travis.yml b/.travis.yml
index fb01bbb7a4f14..4d917a617e0f3 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -35,21 +35,11 @@ env:
 jobs:
   include:
     # Linux environments to build the scikit-learn wheels for the ARM64
-    # architecture and Python 3.7 and newer. This is used both at release time
+    # architecture and Python 3.8 and newer. This is used both at release time
     # with the manual trigger in the commit message in the release branch and as
     # a scheduled task to build the weekly dev build on the main branch. The
     # weekly frequency is meant to avoid depleting the Travis CI credits too
     # fast.
-    - os: linux
-      arch: arm64-graviton2
-      dist: focal
-      virt: vm
-      group: edge
-      if: type = cron or commit_message =~ /\[cd build\]/
-      env:
-        - CIBW_BUILD=cp37-manylinux_aarch64
-        - BUILD_WHEEL=true
-
     - os: linux
       arch: arm64-graviton2
       dist: focal
diff --git a/README.rst b/README.rst
index 0d242a580dddd..a51406ac49b51 100644
--- a/README.rst
+++ b/README.rst
@@ -17,7 +17,7 @@
 .. |Nightly wheels| image:: https://github.com/scikit-learn/scikit-learn/workflows/Wheel%20builder/badge.svg?event=schedule
 .. _`Nightly wheels`: https://github.com/scikit-learn/scikit-learn/actions?query=workflow%3A%22Wheel+builder%22+event%3Aschedule
 
-.. |PythonVersion| image:: https://img.shields.io/badge/python-3.7%20%7C%203.8%20%7C%203.9%20%7C%203.10-blue
+.. |PythonVersion| image:: https://img.shields.io/badge/python-3.8%20%7C%203.9%20%7C%203.10-blue
 .. _PythonVersion: https://pypi.org/project/scikit-learn/
 
 .. |PyPi| image:: https://img.shields.io/pypi/v/scikit-learn
@@ -32,14 +32,14 @@
 .. |Benchmark| image:: https://img.shields.io/badge/Benchmarked%20by-asv-blue
 .. _`Benchmark`: https://scikit-learn.org/scikit-learn-benchmarks/
 
-.. |PythonMinVersion| replace:: 3.7
-.. |NumPyMinVersion| replace:: 1.14.6
-.. |SciPyMinVersion| replace:: 1.1.0
+.. |PythonMinVersion| replace:: 3.8
+.. |NumPyMinVersion| replace:: 1.16.6
+.. |SciPyMinVersion| replace:: 1.3.2
 .. |JoblibMinVersion| replace:: 1.0.0
 .. |ThreadpoolctlMinVersion| replace:: 2.0.0
-.. |MatplotlibMinVersion| replace:: 2.2.3
+.. |MatplotlibMinVersion| replace:: 3.1.2
 .. |Scikit-ImageMinVersion| replace:: 0.14.5
-.. |PandasMinVersion| replace:: 0.25.0
+.. |PandasMinVersion| replace:: 1.0.5
 .. |SeabornMinVersion| replace:: 0.9.0
 .. |PytestMinVersion| replace:: 5.0.1
 
@@ -75,8 +75,8 @@ scikit-learn requires:
 =======
 
 **Scikit-learn 0.20 was the last version to support Python 2.7 and Python 3.4.**
-scikit-learn 0.23 and later require Python 3.6 or newer.
 scikit-learn 1.0 and later require Python 3.7 or newer.
+scikit-learn 1.1 and later require Python 3.8 or newer.
 
 Scikit-learn plotting capabilities (i.e., functions start with ``plot_`` and
 classes end with "Display") require Matplotlib (>= |MatplotlibMinVersion|).
diff --git a/azure-pipelines.yml b/azure-pipelines.yml
index d1251e40e7a4c..42efc0a0c65b9 100644
--- a/azure-pipelines.yml
+++ b/azure-pipelines.yml
@@ -161,10 +161,10 @@ jobs:
         ne(variables['Build.Reason'], 'Schedule')
       )
     matrix:
-      py37_conda_forge_openblas_ubuntu_1804:
+      py38_conda_forge_openblas_ubuntu_1804:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.7'
+        PYTHON_VERSION: '3.8'
         BLAS: 'openblas'
         COVERAGE: 'false'
         BUILD_WITH_ICC: 'false'
@@ -190,11 +190,11 @@ jobs:
         PANDAS_VERSION: 'none'
         THREADPOOLCTL_VERSION: 'min'
         COVERAGE: 'false'
-      # Linux + Python 3.7 build with OpenBLAS
-      py37_conda_defaults_openblas:
+      # Linux + Python 3.8 build with OpenBLAS
+      py38_conda_defaults_openblas:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'defaults'  # Anaconda main channel
-        PYTHON_VERSION: '3.7'
+        PYTHON_VERSION: '3.8'
         BLAS: 'openblas'
         NUMPY_VERSION: 'min'
         SCIPY_VERSION: 'min'
@@ -225,7 +225,7 @@ jobs:
     matrix:
       debian_atlas_32bit:
         DISTRIB: 'debian-32'
-        DOCKER_CONTAINER: 'i386/debian:10.9'
+        DOCKER_CONTAINER: 'i386/debian:11.2'
         JOBLIB_VERSION: 'min'
         # disable pytest xdist due to unknown bug with 32-bit container
         PYTEST_XDIST_VERSION: 'none'
@@ -268,10 +268,10 @@ jobs:
         ne(variables['Build.Reason'], 'Schedule')
       )
     matrix:
-      py37_conda_forge_mkl:
+      py38_conda_forge_mkl:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.7'
+        PYTHON_VERSION: '3.8'
         CHECK_WARNINGS: 'true'
         PYTHON_ARCH: '64'
         # Unpin when pytest stalling issue is fixed
@@ -280,6 +280,6 @@ jobs:
         # Temporary fix for setuptools to use disutils from standard lib
         # https://github.com/numpy/numpy/issues/17216
         SETUPTOOLS_USE_DISTUTILS: 'stdlib'
-      py37_pip_openblas_32bit:
-        PYTHON_VERSION: '3.7'
+      py38_pip_openblas_32bit:
+        PYTHON_VERSION: '3.8'
         PYTHON_ARCH: '32'
diff --git a/build_tools/circle/build_doc.sh b/build_tools/circle/build_doc.sh
index aa932a620d8e1..c0bbb4350cc0a 100755
--- a/build_tools/circle/build_doc.sh
+++ b/build_tools/circle/build_doc.sh
@@ -177,11 +177,6 @@ mamba create -n $CONDA_ENV_NAME --yes --quiet \
     compilers
 
 source activate testenv
-# Pin PyWavelet to 1.1.1 that is the latest version that support our minumum
-# NumPy version required. If PyWavelets 1.2+ is installed, it would require
-# NumPy 1.17+ that trigger a bug with Pandas 0.25:
-# https://github.com/numpy/numpy/issues/18355#issuecomment-774610226
-pip install PyWavelets==1.1.1
 pip install "$(get_dep scikit-image $SCIKIT_IMAGE_VERSION)"
 pip install "$(get_dep sphinx-gallery $SPHINX_GALLERY_VERSION)"
 pip install "$(get_dep numpydoc $NUMPYDOC_VERSION)"
diff --git a/doc/developers/advanced_installation.rst b/doc/developers/advanced_installation.rst
index d89067bc58467..2e8411e264019 100644
--- a/doc/developers/advanced_installation.rst
+++ b/doc/developers/advanced_installation.rst
@@ -67,7 +67,7 @@ feature, code or documentation improvement).
      conda activate sklearn-env
 
 #. **Alternative to conda:** If you run Linux or similar, you can instead use
-   your system's Python provided it is recent enough (3.7 or higher
+   your system's Python provided it is recent enough (3.8 or higher
    at the time of writing). In this case, we recommend to create a dedicated
    virtualenv_ and install the scikit-learn build dependencies with pip:
 
@@ -114,17 +114,12 @@ Runtime dependencies
 Scikit-learn requires the following dependencies both at build time and at
 runtime:
 
-- Python (>= 3.7),
+- Python (>= 3.8),
 - NumPy (>= |NumpyMinVersion|),
 - SciPy (>= |ScipyMinVersion|),
 - Joblib (>= |JoblibMinVersion|),
 - threadpoolctl (>= |ThreadpoolctlMinVersion|).
 
-.. note::
-
-   For running on PyPy, PyPy3-v5.10+, Numpy 1.14.0+, and scipy 1.1.0+
-   are required. For PyPy, only installation instructions with pip apply.
-
 Build dependencies
 ~~~~~~~~~~~~~~~~~~
 
diff --git a/doc/faq.rst b/doc/faq.rst
index 43ef246594de1..65791934ebe9a 100644
--- a/doc/faq.rst
+++ b/doc/faq.rst
@@ -190,10 +190,13 @@ careful choice of algorithms.
 Do you support PyPy?
 --------------------
 
-In case you didn't know, `PyPy <https://pypy.org/>`_ is an alternative
-Python implementation with a built-in just-in-time compiler. Experimental
-support for PyPy3-v5.10+ has been added, which requires Numpy 1.14.0+,
-and scipy 1.1.0+.
+scikit-learn is regularly tested and maintained to work with
+`PyPy <https://pypy.org/>`_ (an alternative Python implementation with
+a built-in just-in-time compiler).
+
+Note however that this support is still considered experimental and specific
+components might behave slightly differently. Please refer to the test
+suite of a the specific module of interest for more details.
 
 How do I deal with string data (or trees, graphs...)?
 -----------------------------------------------------
diff --git a/doc/install.rst b/doc/install.rst
index 2eaf4f299a027..e635790a1962d 100644
--- a/doc/install.rst
+++ b/doc/install.rst
@@ -142,7 +142,8 @@ purpose.
     Scikit-learn 0.21 supported Python 3.5-3.7.
     Scikit-learn 0.22 supported Python 3.5-3.8.
     Scikit-learn 0.23 - 0.24 require Python 3.6 or newer.
-    Scikit-learn 1.0 and later requires Python 3.7 or newer.
+    Scikit-learn 1.0 supported Python 3.7-3.10.
+    Scikit-learn 1.1 and later requires Python 3.8 or newer.
 
 
 .. note::
diff --git a/doc/whats_new/v1.1.rst b/doc/whats_new/v1.1.rst
index b0d36364ec333..bf8145a906a65 100644
--- a/doc/whats_new/v1.1.rst
+++ b/doc/whats_new/v1.1.rst
@@ -15,8 +15,8 @@ Version 1.1.0
 Minimal dependencies
 --------------------
 
-Version 1.1.0 of scikit-learn requires python 3.7+, numpy 1.14.6+ and
-scipy 1.1.0+. Optional minimal dependency is matplotlib 2.2.3+.
+Version 1.1.0 of scikit-learn requires python 3.8+, numpy 1.16.6+ and
+scipy 1.3.2+. Optional minimal dependency is matplotlib 3.1.2+.
 
 Put the changes in their relevant module.
 
diff --git a/pyproject.toml b/pyproject.toml
index 60ce3c28ca3e1..b7c7ed5c0ff71 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -9,12 +9,9 @@ requires = [
     # wheels on PyPI
     #
     # see: https://github.com/scipy/oldest-supported-numpy/blob/master/setup.cfg
-    "oldest-supported-numpy; python_version!='3.7' or platform_machine=='aarch64' or platform_system=='AIX' or platform_python_implementation == 'PyPy'",
+    "oldest-supported-numpy",
 
-    # Override oldest-supported-numpy setting because pandas 0.25.0 requires 1.14.6
-    "numpy==1.14.6; python_version=='3.7' and platform_machine!='aarch64' and platform_system!='AIX' and platform_python_implementation != 'PyPy'",
-
-    "scipy>=1.1.0",
+    "scipy>=1.3.2",
 ]
 
 [tool.black]
diff --git a/setup.py b/setup.py
index 99633008c8dfc..8c1665cd8ba98 100755
--- a/setup.py
+++ b/setup.py
@@ -20,7 +20,7 @@
 try:
     import builtins
 except ImportError:
-    # Python 2 compat: just to be able to declare that Python >=3.7 is needed.
+    # Python 2 compat: just to be able to declare that Python >=3.8 is needed.
     import __builtin__ as builtins
 
 # This is a bit (!) hackish: we are setting a global variable so that the
@@ -156,7 +156,7 @@ def build_extensions(self):
 
 except ImportError:
     # Numpy should not be a dependency just to be able to introspect
-    # that python 3.7 is required.
+    # that python 3.8 is required.
     pass
 
 
@@ -230,6 +230,16 @@ def check_package_status(package, min_version):
 
 
 def setup_package():
+
+    # TODO: Require Python 3.8 for PyPy when PyPy3.8 is ready
+    # https://github.com/conda-forge/conda-forge-pinning-feedstock/issues/2089
+    if platform.python_implementation() == "PyPy":
+        python_requires = ">=3.7"
+        required_python_version = (3, 7)
+    else:
+        python_requires = ">=3.8"
+        required_python_version = (3, 8)
+
     metadata = dict(
         name=DISTNAME,
         maintainer=MAINTAINER,
@@ -255,7 +265,6 @@ def setup_package():
             "Operating System :: Unix",
             "Operating System :: MacOS",
             "Programming Language :: Python :: 3",
-            "Programming Language :: Python :: 3.7",
             "Programming Language :: Python :: 3.8",
             "Programming Language :: Python :: 3.9",
             "Programming Language :: Python :: 3.10",
@@ -263,7 +272,7 @@ def setup_package():
             "Programming Language :: Python :: Implementation :: PyPy",
         ],
         cmdclass=cmdclass,
-        python_requires=">=3.7",
+        python_requires=python_requires,
         install_requires=min_deps.tag_to_packages["install"],
         package_data={"": ["*.pxd"]},
         **extra_setuptools_args,
@@ -282,11 +291,12 @@ def setup_package():
 
         metadata["version"] = VERSION
     else:
-        if sys.version_info < (3, 6):
+        if sys.version_info < required_python_version:
+            required_version = "%d.%d" % required_python_version
             raise RuntimeError(
-                "Scikit-learn requires Python 3.7 or later. The current"
+                "Scikit-learn requires Python %s or later. The current"
                 " Python version is %s installed in %s."
-                % (platform.python_version(), sys.executable)
+                % (required_version, platform.python_version(), sys.executable)
             )
 
         check_package_status("numpy", min_deps.NUMPY_MIN_VERSION)
diff --git a/sklearn/_loss/tests/test_loss.py b/sklearn/_loss/tests/test_loss.py
index 9a56f12ea0039..258bf3ff2ab87 100644
--- a/sklearn/_loss/tests/test_loss.py
+++ b/sklearn/_loss/tests/test_loss.py
@@ -8,6 +8,7 @@
     minimize,
     minimize_scalar,
     newton,
+    LinearConstraint,
 )
 from scipy.special import logsumexp
 
@@ -840,18 +841,13 @@ def fun(x):
     else:
         # The constraint corresponds to sum(raw_prediction) = 0. Without it, we would
         # need to apply loss.symmetrize_raw_prediction to opt.x before comparing.
-        # TODO: With scipy 1.1.0, one could use
-        # LinearConstraint(np.ones((1, loss.n_classes)), 0, 0)
         opt = minimize(
             fun,
             np.zeros((loss.n_classes)),
             tol=1e-13,
             options={"maxiter": 100},
             method="SLSQP",
-            constraints={
-                "type": "eq",
-                "fun": lambda x: np.ones((1, loss.n_classes)) @ x,
-            },
+            constraints=LinearConstraint(np.ones((1, loss.n_classes)), 0, 0),
         )
         grad = loss.gradient(
             y_true=y_true,
diff --git a/sklearn/_min_dependencies.py b/sklearn/_min_dependencies.py
index 5aab7b02efef9..c8c4ad985dcc0 100644
--- a/sklearn/_min_dependencies.py
+++ b/sklearn/_min_dependencies.py
@@ -7,16 +7,9 @@
 if platform.python_implementation() == "PyPy":
     NUMPY_MIN_VERSION = "1.19.0"
 else:
-    # We pinned PyWavelet (a scikit-image dependence) to 1.1.1 in the minimum
-    # documentation CI builds that is the latest version that support our
-    # minimum NumPy version required. If PyWavelets 1.2+ is installed, it would
-    # require NumPy 1.17+ that trigger a bug with Pandas 0.25:
-    # https://github.com/numpy/numpy/issues/18355#issuecomment-774610226
-    # When upgrading NumPy, we can unpin PyWavelets but we need to update the
-    # minimum version of Pandas >= 1.0.5.
-    NUMPY_MIN_VERSION = "1.14.6"
+    NUMPY_MIN_VERSION = "1.16.6"
 
-SCIPY_MIN_VERSION = "1.1.0"
+SCIPY_MIN_VERSION = "1.3.2"
 JOBLIB_MIN_VERSION = "1.0.0"
 THREADPOOLCTL_MIN_VERSION = "2.0.0"
 PYTEST_MIN_VERSION = "5.0.1"
@@ -32,9 +25,9 @@
     "joblib": (JOBLIB_MIN_VERSION, "install"),
     "threadpoolctl": (THREADPOOLCTL_MIN_VERSION, "install"),
     "cython": (CYTHON_MIN_VERSION, "build"),
-    "matplotlib": ("2.2.3", "benchmark, docs, examples, tests"),
+    "matplotlib": ("3.1.2", "benchmark, docs, examples, tests"),
     "scikit-image": ("0.14.5", "docs, examples, tests"),
-    "pandas": ("0.25.0", "benchmark, docs, examples, tests"),
+    "pandas": ("1.0.5", "benchmark, docs, examples, tests"),
     "seaborn": ("0.9.0", "docs, examples"),
     "memory_profiler": ("0.57.0", "benchmark, docs"),
     "pytest": (PYTEST_MIN_VERSION, "tests"),
diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index 68ab834202753..b399f805a9d40 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -20,7 +20,6 @@
 from ..metrics._dist_metrics import METRIC_MAPPING
 from ..utils import check_array
 from ..utils._fast_dict import IntFloatDict
-from ..utils.fixes import _astype_copy_false
 from ..utils.graph import _fix_connected_components
 from ..utils.validation import check_memory
 
@@ -123,7 +122,7 @@ def _single_linkage_tree(
     from scipy.sparse.csgraph import minimum_spanning_tree
 
     # explicitly cast connectivity to ensure safety
-    connectivity = connectivity.astype("float64", **_astype_copy_false(connectivity))
+    connectivity = connectivity.astype(np.float64, copy=False)
 
     # Ensure zero distances aren't ignored by setting them to "epsilon"
     epsilon_value = np.finfo(dtype=connectivity.data.dtype).eps
@@ -559,9 +558,7 @@ def linkage_tree(
     del diag_mask
 
     if affinity == "precomputed":
-        distances = X[connectivity.row, connectivity.col].astype(
-            "float64", **_astype_copy_false(X)
-        )
+        distances = X[connectivity.row, connectivity.col].astype(np.float64, copy=False)
     else:
         # FIXME We compute all the distances, while we could have only computed
         # the "interesting" distances
diff --git a/sklearn/cluster/tests/test_k_means.py b/sklearn/cluster/tests/test_k_means.py
index 8d979bf7f4c48..6afe9bfb8d719 100644
--- a/sklearn/cluster/tests/test_k_means.py
+++ b/sklearn/cluster/tests/test_k_means.py
@@ -9,7 +9,6 @@
 
 from sklearn.utils._testing import assert_array_equal
 from sklearn.utils._testing import assert_allclose
-from sklearn.utils.fixes import _astype_copy_false
 from sklearn.utils.fixes import threadpool_limits
 from sklearn.base import clone
 from sklearn.exceptions import ConvergenceWarning
@@ -783,7 +782,7 @@ def test_float_precision(Estimator, data):
     labels = {}
 
     for dtype in [np.float64, np.float32]:
-        X = data.astype(dtype, **_astype_copy_false(data))
+        X = data.astype(dtype, copy=False)
         km.fit(X)
 
         inertia[dtype] = km.inertia_
diff --git a/sklearn/feature_extraction/text.py b/sklearn/feature_extraction/text.py
index 408ee507a2069..e9c0d14a5eb3b 100644
--- a/sklearn/feature_extraction/text.py
+++ b/sklearn/feature_extraction/text.py
@@ -32,7 +32,6 @@
 from ..utils.validation import check_is_fitted, check_array, FLOAT_DTYPES, check_scalar
 from ..utils.deprecation import deprecated
 from ..utils import _IS_32BIT
-from ..utils.fixes import _astype_copy_false
 from ..exceptions import NotFittedError
 
 
@@ -1621,7 +1620,7 @@ def fit(self, X, y=None):
         if self.use_idf:
             n_samples, n_features = X.shape
             df = _document_frequency(X)
-            df = df.astype(dtype, **_astype_copy_false(df))
+            df = df.astype(dtype, copy=False)
 
             # perform idf smoothing if required
             df += int(self.smooth_idf)
diff --git a/sklearn/feature_selection/_mutual_info.py b/sklearn/feature_selection/_mutual_info.py
index d0dcfa16104b8..47db2601c44c0 100644
--- a/sklearn/feature_selection/_mutual_info.py
+++ b/sklearn/feature_selection/_mutual_info.py
@@ -9,7 +9,6 @@
 from ..neighbors import NearestNeighbors, KDTree
 from ..preprocessing import scale
 from ..utils import check_random_state
-from ..utils.fixes import _astype_copy_false
 from ..utils.validation import check_array, check_X_y
 from ..utils.multiclass import check_classification_targets
 
@@ -287,7 +286,7 @@ def _estimate_mi(
             )
 
         # Add small noise to continuous features as advised in Kraskov et. al.
-        X = X.astype(float, **_astype_copy_false(X))
+        X = X.astype(np.float64, copy=False)
         means = np.maximum(1, np.mean(np.abs(X[:, continuous_mask]), axis=0))
         X[:, continuous_mask] += (
             1e-10
diff --git a/sklearn/linear_model/_coordinate_descent.py b/sklearn/linear_model/_coordinate_descent.py
index f39cbc1e4d990..e47f98e60c94e 100644
--- a/sklearn/linear_model/_coordinate_descent.py
+++ b/sklearn/linear_model/_coordinate_descent.py
@@ -22,7 +22,6 @@
 from ..utils.validation import check_random_state
 from ..model_selection import check_cv
 from ..utils.extmath import safe_sparse_dot
-from ..utils.fixes import _astype_copy_false
 from ..utils.validation import (
     _check_sample_weight,
     check_consistent_length,
@@ -68,9 +67,7 @@ def _set_order(X, y, order="C"):
     if order is not None:
         sparse_format = "csc" if order == "F" else "csr"
         if sparse_X:
-            # As of scipy 1.1.0, new argument copy=False by default.
-            # This is what we want.
-            X = X.asformat(sparse_format, **_astype_copy_false(X))
+            X = X.asformat(sparse_format, copy=False)
         else:
             X = np.asarray(X, order=order)
         if sparse_y:
diff --git a/sklearn/metrics/cluster/_supervised.py b/sklearn/metrics/cluster/_supervised.py
index eb65982801df6..00faab56df169 100644
--- a/sklearn/metrics/cluster/_supervised.py
+++ b/sklearn/metrics/cluster/_supervised.py
@@ -23,7 +23,6 @@
 from scipy import sparse as sp
 
 from ._expected_mutual_info_fast import expected_mutual_information
-from ...utils.fixes import _astype_copy_false
 from ...utils.multiclass import type_of_target
 from ...utils.validation import check_array, check_consistent_length
 
@@ -919,7 +918,7 @@ def adjusted_mutual_info_score(
     ):
         return 1.0
     contingency = contingency_matrix(labels_true, labels_pred, sparse=True)
-    contingency = contingency.astype(np.float64, **_astype_copy_false(contingency))
+    contingency = contingency.astype(np.float64, copy=False)
     # Calculate the MI for the two clusterings
     mi = mutual_info_score(labels_true, labels_pred, contingency=contingency)
     # Calculate the expected value for the mutual information
@@ -1029,7 +1028,7 @@ def normalized_mutual_info_score(
     ):
         return 1.0
     contingency = contingency_matrix(labels_true, labels_pred, sparse=True)
-    contingency = contingency.astype(np.float64, **_astype_copy_false(contingency))
+    contingency = contingency.astype(np.float64, copy=False)
     # Calculate the MI for the two clusterings
     mi = mutual_info_score(labels_true, labels_pred, contingency=contingency)
     # Calculate the expected value for the mutual information
@@ -1113,7 +1112,7 @@ def fowlkes_mallows_score(labels_true, labels_pred, *, sparse=False):
     (n_samples,) = labels_true.shape
 
     c = contingency_matrix(labels_true, labels_pred, sparse=True)
-    c = c.astype(np.int64, **_astype_copy_false(c))
+    c = c.astype(np.int64, copy=False)
     tk = np.dot(c.data, c.data) - n_samples
     pk = np.sum(np.asarray(c.sum(axis=0)).ravel() ** 2) - n_samples
     qk = np.sum(np.asarray(c.sum(axis=1)).ravel() ** 2) - n_samples
diff --git a/sklearn/utils/fixes.py b/sklearn/utils/fixes.py
index 34391847eb46b..c34ffd443ad9a 100644
--- a/sklearn/utils/fixes.py
+++ b/sklearn/utils/fixes.py
@@ -15,7 +15,6 @@
 
 import sklearn
 import numpy as np
-import scipy.sparse as sp
 import scipy
 import scipy.stats
 from scipy.sparse.linalg import lsqr as sparse_lsqr  # noqa
@@ -46,18 +45,6 @@ def _object_dtype_isnan(X):
     return X != X
 
 
-# TODO: replace by copy=False, when only scipy > 1.1 is supported.
-def _astype_copy_false(X):
-    """Returns the copy=False parameter for
-    {ndarray, csr_matrix, csc_matrix}.astype when possible,
-    otherwise don't specify
-    """
-    if sp_version >= parse_version("1.1") or not sp.issparse(X):
-        return {"copy": False}
-    else:
-        return {}
-
-
 class loguniform(scipy.stats.reciprocal):
     """A class supporting log-uniform random variables.